framework,version,device,op_name,kernel_source,gemm_dtype,m,n,k,latency
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,16384,41.35464138454861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,12288,30.625057644314236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,51200,139.43287489149307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,16384,20.667266845703125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,12288,15.6836429172092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,10240,27.045591566297745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,12288,61.278401692708336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,51200,63.36512586805555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,10240,12.829583909776476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,16384,84.48388671875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,8192,26.402413262261287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,8192,10.505620320638021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,10240,51.044219970703125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,7168,22.953082614474827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,7168,9.09779781765408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,7168,35.65710110134548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,8192,43.693935818142364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,6144,19.324966430664062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,6144,8.417351616753473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,6144,30.720231797960068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,5120,16.130321926540798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,5120,6.90994135538737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,4096,12.780724419487846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,8192,29.549479166666668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,5120,25.711841159396702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,4096,15.024112277560763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,4096,5.76419915093316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,6144,23.07393731011285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,3584,10.225381639268663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,7168,26.85915968153212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,3584,4.634880065917969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,3584,13.604649861653646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,3584,18.032340155707466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,3072,8.884745279947916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,3072,4.120417700873481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,4096,20.61578199598524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,2560,3.6522988213433156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,3072,13.764642503526476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,2560,7.557612948947483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,2560,8.82003868950738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,2560,8.485887315538195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,2048,2.999728944566515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,2048,6.1071929931640625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,2048,6.458228641086155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,1536,2.1396772596571183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,1536,4.61863538953993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,1536,4.580816904703776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,1536,4.662386576334636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,1024,3.245603561401367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,1024,1.4238221910264757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,1024,3.2455395592583547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,1024,3.3590142991807728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,768,2.690928988986545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,768,1.165628433227539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,768,2.5361830393473306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,768,2.589883592393663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,512,1.8969039916992188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,512,0.980412483215332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,512,1.9410720401340062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,512,2.024295171101888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,256,1.3134720060560439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,256,0.8559981981913248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,128,0.8635226885477701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,256,1.7504747178819444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,128,0.8297244707743326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,256,1.4413813485039606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,128,1.656221283806695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,64,0.7002595795525445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,64,1.6916764577229817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,32,2.1722933451334634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,32,0.7159200244479709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,51200,253.8905707465278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,5120,18.670401679144966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,2048,6.094543880886501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,128,1.0822622511121962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,65536,63.243069118923614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,51200,49.329742431640625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,51200,110.42835828993056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,65536,138.72706434461807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,16384,32.51652357313368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,16384,15.64384290907118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,16384,45.991255018446175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,12288,24.105173746744793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,16384,66.87664116753471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,12288,11.993337843153212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,12288,34.56131320529514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,12288,48.05730183919271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,10240,9.988248189290365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,10240,27.900066799587673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,51200,198.6124945746528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,8192,16.40869310167101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,10240,28.990237765842014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,8192,8.070793151855469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,10240,39.844685872395836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,8192,34.368489583333336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,7168,17.846661037868923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,8192,24.05007765028212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,7168,7.175142076280381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,7168,27.943596733940975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,6144,15.535910712348091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,6144,6.312264760335286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,6144,17.541778564453125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,6144,23.995142618815105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,65536,192.24217393663196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,5120,12.776876661512587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,5120,5.591896057128906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,65536,278.41425238715277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,4096,9.647709316677517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,4096,4.420286390516493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,5120,14.59846666124132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,5120,20.142895168728298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,4096,11.871278550889757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,4096,16.11937798394097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,3584,3.750402662489149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,51200,147.31184895833334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,3584,8.561543782552084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,3072,3.3349901835123696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,3072,6.776967790391709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,3584,10.302495320638021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,3584,14.114266289605034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,2560,6.3133290608723955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,2560,2.6478195190429688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,3072,8.697035047743055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,3072,10.880269368489584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,2560,6.6359405517578125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,2048,1.9777092403835719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,2560,6.525850507948134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,2048,5.408991919623481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,2048,4.8651652865939665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,2048,4.85456805759006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,1536,1.5202897389729817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,7168,20.403771294487846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,1536,3.6510594685872397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,1536,4.191786872016059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,1024,2.392952813042535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,1536,3.604523552788628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,1024,2.5938267178005643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,1024,2.6061119503445096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,768,1.8336274888780382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,1024,1.0151920318603516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,768,0.9731261995103625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,512,1.4887751473320856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,768,1.9585325453016493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,512,0.7909457948472766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,256,0.6707626448737251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,512,1.5651182598537867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,128,0.6962266498141818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,512,1.521359125773112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,256,1.1681280136108398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,128,0.6664382086859809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,64,0.5390702353583442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,128,1.2955901887681749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,32,0.5625991291469997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,256,0.9028693305121528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,64,1.3229920069376628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,256,1.380708482530382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,32,1.6992337968614366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,65536,20.288991292317707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,51200,35.125267876519096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,65536,42.32838609483507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,51200,16.624242146809895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,16384,11.678520202636719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,51200,37.60978529188368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,128,0.8484435611300998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,768,2.0365573035346136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,16384,5.178236643473308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,16384,12.447205437554253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,51200,65.13471815321181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,12288,3.957861158582899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,65536,49.60551961263021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,12288,9.09950425889757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,10240,3.1641065809461804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,16384,21.502010769314236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,10240,8.235281202528212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,12288,9.194815741644964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,8192,2.124463187323676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,12288,15.569936116536459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,10240,7.833648681640625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,8192,6.276702033148871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,8192,6.344800737169053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,7168,1.8095618353949652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,7168,5.21049329969618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,10240,13.091190761990019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,7168,5.397377014160156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,6144,1.600129763285319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,6144,4.829438951280382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,8192,10.940105862087675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,7168,8.953928629557291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,6144,4.555931515163846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,5120,3.538840823703342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,5120,1.182903077867296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,6144,6.1557464599609375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,4096,0.9276373121473523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,5120,4.007169087727864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,4096,3.0966292487250437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,5120,3.662065717909071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,4096,2.981320063273112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,4096,2.8178488413492837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,3584,0.8806062274509006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,3584,2.6282816992865667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,3584,2.3788320753309464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,3584,2.5420479244656033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,3072,0.7778328789605035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,3072,2.243506749471029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,3072,2.0642674763997397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,65536,88.85540771484375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,2560,0.6304604212443033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,3072,2.1310515933566623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,2560,1.996978759765625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,2560,1.6948187086317275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,2048,0.5267857975429958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,2048,1.4456550810072157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,2560,1.7895103030734594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,1536,1.066197289360894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,2048,1.3369457456800673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,1024,0.8057004610697428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,2048,1.394800927903917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,1536,1.0768719779120552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,1024,0.33160000377231175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,1536,1.0868284437391493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,768,0.6044106483459473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,768,0.29072178734673393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,1024,0.7753680017259387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,1536,0.4514924685160319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,768,0.628760019938151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,768,0.599273787604438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,256,0.30139022403293186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,512,0.26820090081956655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,512,0.5534817907545302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,128,0.21590399742126465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,256,0.23495644993252227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,512,0.4761502477857802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,256,0.37869776619805234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,64,0.17313244607713488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,128,0.2344364325205485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,128,0.423110220167372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,32,0.1811279985639784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,64,0.4247431225246853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,1024,0.7854800224304199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,32,0.4798933135138617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,256,0.444834656185574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,512,0.4316062132517497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,65536,15.673873901367188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,51200,24.74514940049913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,51200,12.307892693413628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,51200,29.377166748046875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,16384,8.590743170844183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,65536,35.09785630967882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,16384,3.980897691514757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,16384,9.478045993381077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,12288,6.832124498155381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,12288,2.4740382300482855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,16384,16.403333875868054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,12288,7.094730801052517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,51200,49.45007663302951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,10240,5.566351996527778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,12288,11.788909912109375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,10240,2.0151360829671225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,8192,4.1428070068359375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,10240,5.959612528483073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,10240,9.837320115831163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,8192,1.493742201063368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,8192,4.549254947238498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,8192,6.148311191134983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,65536,37.72174072265625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,7168,1.3047644297281902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,7168,3.9386435614691844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,6144,1.0454097323947482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,7168,4.61665513780382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,7168,3.8312899271647134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,6144,3.6530418395996094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,6144,3.3701243930392795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,6144,3.1806488037109375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,5120,0.9244195090399848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,5120,3.0359590318467884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,5120,2.612414254082574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,4096,2.256792916191949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,4096,0.7570408715142144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,128,0.2891457875569661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,4096,2.0097423129611545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,5120,2.7290933397081165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,4096,2.034887949625651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,3584,0.6255911191304525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,3584,2.1362747616238065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,3584,1.7724640104505751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,3072,0.548351976606581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,3072,1.8288951449924047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,3584,1.825064977010091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,3072,1.5393750932481554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,2560,0.4880337715148926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,2560,1.2903430726793077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,3072,1.4804133309258356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,2560,1.267579608493381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,2048,0.40671645270453555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,2048,1.031225734286838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,2560,1.6244204839070637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,1536,0.846127986907959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,2048,1.0499484803941515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,1536,0.8246542082892524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,1536,0.3276924557156033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,1536,0.7857635815938314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,1024,0.6141680081685384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,1024,0.2550186581081814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,1024,0.603500419192844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,768,0.4313155545128717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,1024,0.5662364429897732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,768,0.23251554701063368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,512,0.3378293249342177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,2048,1.194799953036838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,768,0.4555075433519151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,512,0.3909200032552083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,256,0.23468266593085396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,512,0.26580622461107045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,512,0.36320800251430935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,128,0.2259653409322103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,256,0.3379955556657579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,64,0.1314364406797621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,128,0.3197644551595052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,768,0.5567279921637641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,32,0.1377484401067098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,64,0.3199084334903293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,32,0.35047290060255265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,256,0.18407911724514434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,65536,67.14133029513889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,128,0.1929804484049479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,256,0.28579288058810765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,65536,13.568477206759981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,51200,20.38309563530816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,51200,10.631617228190104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,51200,24.801600138346355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,16384,6.938779619004991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,16384,3.1994071536593967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,65536,33.04181586371528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,16384,7.896912468804254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,12288,5.221622043185763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,12288,2.0088844299316406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,16384,14.044712490505644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,51200,41.03999498155382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,12288,9.740553114149305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,12288,5.915435367160374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,10240,4.610196431477864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,10240,1.5721297793918187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,8192,3.662316640218099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,10240,4.865125444200304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,10240,7.747571309407552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,8192,3.6023402743869357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,8192,1.1618888643052843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,8192,4.55207273695204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,7168,0.9798790613810221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,7168,3.1905004713270397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,7168,3.1438204447428384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,65536,31.38046603732639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,7168,3.0434898800320096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,6144,3.1517990960015188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,6144,0.8623510996500651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,6144,2.9043057759602866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,5120,0.8194106419881185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,6144,2.611944834391276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,5120,2.12773068745931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,5120,2.181362575954861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,5120,2.098049799601237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,4096,2.1544134351942277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,4096,0.7164017889234754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,4096,1.7434284422132704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,4096,1.7100266350640192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,3584,0.5312844382392036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,3584,1.6894729402330186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,3584,1.5181973775227864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,3072,1.491777737935384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,3584,1.5065519544813368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,3072,1.3542683919270833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,3072,1.281820403205024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,2560,0.4128906726837158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,3072,0.48659997516208225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,2560,1.1166071361965604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,2560,1.0430035061306424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,2048,0.9250906838311089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,2048,0.36029688517252606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,2048,0.8895698123508029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,2560,1.2770871056450737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,2048,0.8543342484368218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,1536,0.28135376506381565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,1024,0.45776621500651044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,1536,0.7971564398871528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,1536,0.6530808872646755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,1024,0.2531324492560493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,768,0.3550533453623454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,1536,0.6446631219651964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,1024,0.4786426756117079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,768,0.21873777442508271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,768,0.4066693252987332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,512,0.271724436018202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,768,0.3853911028967963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,256,0.18249332904815674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,512,0.3214640087551541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,512,0.3089582125345866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,256,0.2848079999287923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,256,0.2412355475955539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,128,0.13261244032118055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,512,0.17922488848368326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,128,0.2680853207906087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,64,0.11300444602966309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,1024,0.5231759813096788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,32,0.118212448226081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,256,0.16153155432807073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,64,0.2693866623772515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,32,0.2853226661682129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,65536,56.045196533203125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,128,0.16254044903649226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,128,0.2207946644888984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,65536,10.793971591525606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,51200,15.991217719184029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,51200,8.316969977484808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,65536,21.205642700195312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,16384,5.994927300347222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,51200,19.87597147623698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,16384,2.2096773783365884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,16384,6.2837066650390625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,12288,4.370879279242621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,12288,1.5305386649237738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,16384,10.711881849500868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,12288,7.248200310601129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,12288,4.672368791368273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,51200,31.976318359375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,10240,3.9631377326117625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,10240,1.1770132912529838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,10240,4.272095150417752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,10240,3.7583385043674045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,8192,3.0983520083957248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,8192,0.9612639745076498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,8192,3.202566146850586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,8192,2.822896957397461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,7168,0.840519110361735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,7168,2.7979920705159507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,6144,2.119661331176758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,7168,2.539632797241211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,7168,2.441127141316732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,6144,0.7192239761352539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,6144,2.2429786258273654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,5120,1.8586124844021266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,5120,0.5997582011752659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,5120,1.8500542110866969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,6144,2.077256096733941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,4096,0.49159823523627383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,4096,1.471660402086046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,4096,1.340695063273112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,4096,1.3329422208997939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,5120,1.7172444661458333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,3584,0.43141510751512313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,3584,1.2052720387776692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,3584,1.1251271565755208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,3072,1.1135742399427626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,3584,1.2802711062961154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,3072,1.0709173414442275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,3072,0.9975013732910156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,2560,0.9725075827704536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,2560,0.8735235532124838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,65536,25.517393324110245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,2560,0.34241512086656356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,3072,0.39689156744215226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,2048,0.7841777801513672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,2048,0.2909324434068468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,2048,0.7143155733744303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,1536,0.5431306627061632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,1536,0.238663117090861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,1024,0.3576337761349148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,2048,0.6793200174967448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,2560,0.8233493169148763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,1024,0.19390488995446098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,768,0.27477778328789604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,1024,0.4077484342787001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,1024,0.4330320093366835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,768,0.17029333114624023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,768,0.3199884361690945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,512,0.21140621768103707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,768,0.31824710634019643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,1536,0.5193635622660319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,256,0.14950666162702772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,512,0.26359465387132436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,512,0.25099023183186847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,65536,44.036098904079864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,256,0.1378151045905219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,256,0.2315973440806071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,256,0.1953484482235379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,64,0.08827200200822617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,128,0.13896355364057753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,128,0.2171351114908854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,32,0.09296177493201362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,1536,0.7997440232170953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,64,0.2183004485236274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,32,0.22547290060255262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,128,0.10706666443083022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,128,0.15238489045037165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,512,0.1600088940726386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,65536,9.980519612630209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,51200,14.383171929253471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,128,0.18663199742635092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,65536,18.71335177951389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,51200,17.388154771592884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,16384,4.499847835964627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,16384,1.844130622016059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,51200,7.877597384982639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,12288,3.3822506798638234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,12288,1.3174595303005643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,16384,5.553026835123698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,16384,8.592715793185764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,12288,5.029056972927518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,10240,3.3254871368408203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,10240,3.373475604587131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,10240,3.1444924672444663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,10240,1.1033155653211806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,51200,28.665601942274307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,8192,0.8353581958346896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,12288,3.916102091471354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,8192,2.4119288126627603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,8192,2.467834684583876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,7168,0.7147910859849719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,7168,2.3074177636040583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,7168,2.0948284996880426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,7168,2.121084425184462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,8192,2.574357350667318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,6144,0.6226515769958496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,6144,2.2039093441433377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,6144,1.950076421101888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,5120,1.6756728490193684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,5120,0.5539911058213975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,5120,1.5223911073472765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,5120,1.4912346733940973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,4096,0.4360106786092122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,6144,1.8348426818847656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,4096,1.2511279847886827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,4096,1.202968915303548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,3584,1.0312301847669814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,3584,0.5165368715922037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,3584,1.0937777625189888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,3584,1.0445920096503365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,3072,0.9343173768785266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,3072,0.34708889325459796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,4096,1.1606488757663302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,3072,0.8578053580390083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,2560,0.302818669213189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,2560,0.8649528821309408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,2048,0.5930079883999294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,2560,0.7248515552944608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,2048,0.3451484574211969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,3072,1.019568019443088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,2048,0.6269022093878852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,2048,0.5801484319898818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,1536,0.2085662285486857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,1024,0.32847823037041557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,1536,0.5104524294535319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,1536,0.4606453577677409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,1024,0.16750310526953804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,768,0.2518533335791694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,2560,0.8392933209737142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,1024,0.3745271099938287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,768,0.15527289443545872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,768,0.28280356195237905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,768,0.2794613308376736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,512,0.23324177000257704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,512,0.22239910231696236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,256,0.12661333878835043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,256,0.2040728860431247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,512,0.17458044158087838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,256,0.17286400000254312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,128,0.09607466724183823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,256,0.12650222248501247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,512,0.14026489522722033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,128,0.19076800346374512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,64,0.08300177918540107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,65536,22.153978135850693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,128,0.13537600305345324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,32,0.08762400017844306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,64,0.19165866904788545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,32,0.24446490075853136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,128,0.126601775487264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,1024,0.358534230126275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,65536,38.61388821072048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,1536,0.42206843694051105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,51200,6.432403564453125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,65536,8.324843512641058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,51200,12.219215393066406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,16384,4.122204250759549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,16384,1.4849653244018555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,16384,6.600474463568793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,51200,15.19663831922743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,65536,17.917809380425346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,16384,4.6708496941460504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,12288,3.059867646959093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,12288,1.1118915345933702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,12288,3.8782632615831165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,10240,2.9868062337239585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,12288,3.3670614030626087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,10240,0.9428204430474175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,10240,2.704200956556532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,10240,2.9717964596218533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,8192,0.7050835291544596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,8192,2.1164071824815536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,8192,2.141496022542318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,8192,2.449397404988607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,7168,0.6278284390767416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,7168,1.9831280178493922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,7168,1.7944791581895616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,51200,24.8539055718316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,6144,0.5436222288343642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,6144,1.6104052861531575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,5120,1.3268586264716256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,6144,1.6234027014838324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,6144,1.470255109998915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,7168,1.8879280090332031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,5120,0.6517857975429958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,5120,1.3680773841010199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,5120,1.2271689309014213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,4096,1.0049208535088432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,4096,0.377895991007487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,4096,1.0580497317843967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,3584,0.33953600459628636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,4096,0.9903235965304904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,3584,0.9200222227308484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,3072,0.30566755930582684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,3584,0.8369342486063639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,3072,0.822822199927436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,3072,0.725730684068468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,3584,0.9921716054280599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,2560,0.5936062071058485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,2560,0.2666177749633789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,2560,0.668327119615343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,3072,0.7704977989196777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,2048,0.2281271086798774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,2048,0.6370879809061686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,1536,0.41214932335747617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,2048,0.5033822059631348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,1024,0.2764328850640191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,2048,0.5560577710469564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,1536,0.3998400105370416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,2560,0.6131288740370009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,1024,0.31259022818671334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,768,0.21040711137983534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,1024,0.29888712035285103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,768,0.25065334637959796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,768,0.24402666091918945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,65536,32.208275689019096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,1024,0.15329955683814153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,512,0.20270489321814644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,512,0.19360355536142984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,256,0.10683200094434951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,1536,0.23697511355082193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,256,0.1766479942533705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,256,0.14975999461279974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,128,0.08265422450171576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,128,0.16495022508833143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,128,0.11847821871439616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,512,0.12724978393978542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,64,0.06907111406326294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,128,0.11439200242360432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,32,0.07157599925994873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,512,0.16316978136698404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,64,0.16532088650597468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,256,0.1138862238989936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,32,0.17019377814398873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,768,0.14065333207448324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,1536,0.4280364513397217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,65536,19.438744439019096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,65536,6.928551144070095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,51200,5.207542419433594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,51200,10.59897189670139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,16384,3.40814569261339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,16384,1.2306231392754448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,16384,4.674625820583768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,16384,3.8567093743218313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,12288,2.66518317328559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,65536,14.392032199435763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,51200,12.959278530544706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,12288,0.9476906458536783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,65536,16.629789564344616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,12288,2.9307901594373913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,10240,2.3325352138943143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,12288,2.7629032135009766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,10240,0.8356515566507975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,10240,2.4423804812961154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,10240,2.2219857109917536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,8192,0.5946817927890354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,51200,18.593407524956596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,8192,1.7740738126966689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,7168,0.5373244285583496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,8192,1.9722444746229384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,7168,1.6708621978759766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,7168,1.529004414876302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,8192,1.7619537777370875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,6144,1.3675048616197374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,6144,0.46198309792412656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,6144,1.4319857491387262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,6144,1.288482666015625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,5120,1.1061173544989693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,4096,0.8446684413486056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,5120,1.0356551276312935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,4096,0.9105119705200195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,4096,0.33009688059488934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,4096,0.8616337776184082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,7168,1.8176754845513239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,3584,0.7431422339545356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,3584,0.29827022552490234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,3584,0.7958799997965494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,5120,0.408095121383667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,3072,0.7078328662448459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,3584,0.7533840073479546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,3072,0.26452710893419057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,65536,25.7562255859375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,3072,0.6388586892022027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,2560,0.23134401109483507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,2048,0.42288533846537274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,2560,0.5265795389811198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,2048,0.2038880056805081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,3072,0.677171548207601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,2048,0.4639964633517795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,2560,0.6510657734341091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,1536,0.32620443238152397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,1536,0.16194221708509657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,1536,0.3650373352898492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,1536,0.34045865800645614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,1024,0.13324622313181558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,768,0.18464799722035727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,1024,0.2569173441992866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,768,0.2112728887134128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,768,0.12713689274258083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,768,0.21470222208234999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,2560,0.5874586635165745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,512,0.11330044269561768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,512,0.17318577236599395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,512,0.16572177410125732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,256,0.14959022733900282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,2048,0.4571724467807346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,256,0.12765244642893472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,128,0.07023111316892836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,256,0.10259289211697048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,128,0.10210488902197944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,128,0.13899377981821695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,128,0.10157955355114406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,256,0.09255822499593098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,32,0.06538755363888211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,64,0.1389591164059109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,32,0.14128177695804173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,512,0.14141866895887586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,5120,1.069702254401313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,1024,0.3067537678612603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,64,0.061436441209581166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,1024,0.23234221670362684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,65536,4.923289828830295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,51200,3.8877796596950955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,51200,8.372425503200954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,16384,3.3752320607503257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,65536,10.26820543077257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,16384,1.0415822135077584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,16384,2.7200809054904513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,16384,3.0791929033067493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,12288,2.1272665659586587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,51200,10.176690843370226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,12288,0.732951111263699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,12288,2.4064222971598306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,12288,2.208213382297092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,10240,1.8619217342800563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,10240,0.6503662003411187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,8192,1.465952025519477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,8192,1.5268916024102106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,8192,0.5128924581739638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,51200,15.583034939236112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,10240,1.8372008005777996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,7168,1.2495750851101346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,7168,0.44481600655449766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,8192,1.468804465399848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,7168,1.182776027255588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,7168,1.2791030671861436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,10240,1.8523936801486547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,6144,0.38635823461744523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,6144,1.122479968600803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,6144,1.0627333323160808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,5120,0.8713502354092069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,5120,0.3319057888454861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,5120,0.9096000459459094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,4096,0.6769990921020508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,6144,1.0243999693128798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,5120,0.8728462325202094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,4096,0.2766417662302653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,4096,0.812020460764567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,3584,0.6917040083143445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,3584,0.2499173217349582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,3584,0.6427600118849013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,3072,0.47029866112603086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,3584,0.5945306883917915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,65536,20.174786037868923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,3072,0.2222062216864692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,2560,0.4224079979790582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,3072,0.6252737575107151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,3072,0.5170995394388834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,2560,0.48536353641086155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,2048,0.3671457767486572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,2560,0.43549511167738175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,2048,0.2159351110458374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,1536,0.25580355856153697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,2048,0.3606719970703125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,1536,0.1417271163728502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,2560,0.19570222165849474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,1536,0.35277332199944395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,1536,0.28216621610853404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,4096,0.6718871328565809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,768,0.14097155465020075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,1024,0.11712888876597087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,1024,0.2210106717215644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,1024,0.21222133106655547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,512,0.10346666971842448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,768,0.17515288458930123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,768,0.1724568870332506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,1024,0.1847226619720459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,512,0.13766133785247803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,2048,0.3914622200859918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,256,0.12261333730485703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,768,0.11177333196004231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,128,0.056141330136193164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,256,0.10519644286897446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,128,0.08964266379674275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,512,0.14342221948835585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,64,0.047351999415291675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,256,0.06751466459698148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,32,0.04873155554135641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,64,0.11251733038160537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,32,0.11355288823445638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,256,0.09078222513198853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,128,0.11325244108835857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,512,0.10065955585903591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,65536,4.376302931043837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,65536,13.242971632215712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,51200,7.482460021972656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,51200,3.5438230302598743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,16384,2.3728533850775824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,65536,11.330034891764322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,65536,11.431383768717447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,16384,1.0183093812730577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,16384,2.8465796576605906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,12288,2.1388933393690324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,51200,8.925630357530382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,12288,2.006559160020616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,65536,15.006639268663195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,12288,0.696572462717692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,51200,13.28486802842882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,12288,1.9603004455566406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,10240,0.6360488997565376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,16384,2.774782180786133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,10240,1.6630178027682836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,8192,1.2907164891560872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,8192,0.44969866010877824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,8192,1.3986772961086695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,8192,1.2773528628879123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,7168,1.0892249213324654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,10240,1.6038551330566406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,7168,1.0732942157321506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,7168,0.41018133693271214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,6144,0.9328781763712565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,6144,0.3465146753523085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,6144,1.049581315782335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,5120,0.7428888744778104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,7168,1.144400914510091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,6144,0.8949208789401584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,5120,0.30299467510647243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,5120,0.8476790852016873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,5120,0.7534809112548828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,10240,2.0405466291639542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,4096,0.6511066754659017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,4096,0.6166177855597602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,3584,0.22776532173156738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,3584,0.5888364579942491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,3584,0.5403795772128636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,4096,0.6343386438157823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,4096,0.2514622211456299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,3072,0.20911733309427896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,3072,0.4993884298536513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,2560,0.40273687574598527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,3072,0.4613857799106174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,2560,0.17938666873508027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,2560,0.4256142245398627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,2560,0.4014115598466661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,3072,0.45502487818400067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,2048,0.15419822269015843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,1536,0.2384702099694146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,2048,0.32188354598151314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,1536,0.14885955386691624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,1536,0.2719004419114855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,3584,0.5060604413350424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,1536,0.25403555234273273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,768,0.12886399692959255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,1024,0.20462489128112793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,1024,0.19050133228302002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,2048,0.3485199875301785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,768,0.15785600079430476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,768,0.15502222379048666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,1024,0.15187199910481772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,512,0.12861422697703043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,2048,0.29428354899088544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,512,0.12278489271799724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,512,0.09393244319491917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,512,0.08938488695356582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,256,0.09398844507005479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,256,0.08499555455313788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,128,0.05477866530418396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,1024,0.1099768877029419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,128,0.10033866431978013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,256,0.0683271090189616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,64,0.04663377669122484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,128,0.08369155724843343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,32,0.04859911070929634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,768,0.10501244333055283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,64,0.09911821948157416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,32,0.09990844461652969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,256,0.10883288913302952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,65536,3.704938676622179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,51200,7.102372487386067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,65536,8.139025370279947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,16384,2.199659559461806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,65536,9.941156175401476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,51200,3.1538079579671225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,16384,0.7853777673509387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,51200,7.743964301215278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,16384,2.387151082356771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,16384,2.383972379896376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,51200,12.008212619357638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,65536,13.637749565972221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,12288,0.6072790887620714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,12288,1.8920133378770616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,10240,1.6689786911010742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,12288,1.6932818094889324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,10240,0.5685155656602647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,10240,1.508703973558214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,10240,1.3999564912584093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,8192,0.4098924530877008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,8192,1.2302079730563693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,8192,1.1611821916368272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,8192,1.0796818203396268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,7168,0.3589626683129205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,12288,1.7828114827473958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,7168,1.048160023159451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,7168,0.9846009148491753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,6144,0.9499235153198242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,7168,0.9814675649007162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,6144,0.8948373794555664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,5120,0.6313786506652832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,6144,0.8001599841647677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,5120,0.7295840051439074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,5120,0.26843733257717556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,4096,0.5286995569864908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,5120,0.6643742455376519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,4096,0.22392355071173775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,4096,0.5800791316562229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,3584,0.5102497736612955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,4096,0.5333342022365993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,3584,0.20271200603908965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,3584,0.5268097983466254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,3072,0.3965715567270915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,3584,0.4674968719482422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,3072,0.18553333812289766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,6144,0.3166782326168484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,2560,0.35381245613098145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,3072,0.4087119897206624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,2560,0.16337066226535374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,2560,0.380220439698961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,2560,0.34939556651645237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,2048,0.3052693208058675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,1536,0.20487111144595674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,2048,0.2863386736975776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,3072,0.4419155650668674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,1024,0.13694489002227783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,1536,0.22799910439385307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,1024,0.17541689342922637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,2048,0.14129777749379477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,1024,0.1706035534540812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,1536,0.2402799924214681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,768,0.1393457783593072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,2048,0.2760897742377387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,768,0.09670044316185845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,768,0.13860266738467747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,1024,0.10043111112382676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,512,0.10975822475221421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,1536,0.1189546717537774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,512,0.08684711323844062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,256,0.09546222289403279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,256,0.07858755853441027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,768,0.11134133074018691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,256,0.08244178030225965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,512,0.08685688840018378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,256,0.05446666810247633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,128,0.08715911044014825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,64,0.037696888049443565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,128,0.07760622104008992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,32,0.039015110996034406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,64,0.0856177806854248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,32,0.08630400233798557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,512,0.11308622360229492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,128,0.04357066750526428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,128,0.06717333528730604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,65536,3.2762862311469187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,51200,6.149107615152995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,65536,8.17233615451389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,128,0.08452888992097642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,16384,1.8773839738633897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,51200,2.7095786200629344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,16384,2.1320249769422746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,16384,0.830665800306532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,65536,8.493052164713541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,65536,12.089714050292969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,16384,2.0054310692681208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,51200,6.562619527180989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,51200,9.481383429633246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,12288,0.5391155348883735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,12288,1.6022408803304036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,12288,1.405445310804579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,10240,0.4679475360446506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,12288,1.4490320417616103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,10240,1.253710216946072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,10240,1.2046239640977647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,8192,0.35426489512125653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,8192,1.0230355792575412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,10240,1.1634871164957683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,8192,0.948139508565267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,7168,0.8104773097568088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,7168,0.3153955671522352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,6144,0.6120808919270834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,8192,0.9720746146308051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,6144,0.28446221351623535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,7168,0.8002639876471626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,6144,0.7667662302652994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,5120,0.5202666918436686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,5120,0.23562754525078666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,6144,0.995605362786187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,4096,0.4047439893086751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,5120,0.6383840243021647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,5120,0.5840426550971137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,4096,0.2002577781677246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,3584,0.3662613232930501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,4096,0.5535377926296657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,4096,0.4664870897928874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,3584,0.18208089139726427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,3584,0.45108535554673934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,3584,0.417921781539917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,2560,0.2716533342997233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,3072,0.3565688927968343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,7168,0.9011920293172201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,2560,0.32907289928860134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,2560,0.14491021633148193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,2560,0.3068702220916748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,3072,0.3313608964284261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,2048,0.1279217799504598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,1536,0.1624764468934801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,3072,0.1639031171798706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,3072,0.38799556096394855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,2048,0.2500142256418864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,1024,0.11838400363922119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,1536,0.10805422067642212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,1024,0.09186489052242702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,1024,0.15247378084394667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,1024,0.15005332893795437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,2048,0.2666177749633789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,768,0.12120178010728623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,512,0.06614222129185994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,768,0.0896586643324958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,512,0.0981866651111179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,1536,0.19666755199432373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,512,0.094449778397878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,256,0.04734488990571764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,512,0.0799822211265564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,256,0.07270666625764635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,256,0.08191733227835761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,128,0.03722666700681051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,256,0.07112355364693536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,128,0.07166577710045709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,1536,0.21026933193206787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,768,0.0958906677034166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,32,0.037131554550594754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,64,0.07298044363657634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,768,0.12087466981675889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,32,0.07225066423416138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,128,0.07435644335216947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,2048,0.21138132943047416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,64,0.03532977898915609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,128,0.05887111028035482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,65536,2.768705791897244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,51200,4.653000725640191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,65536,6.333213382297092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,51200,2.2714551289876304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,16384,1.4089973237779405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,65536,7.037499745686849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,16384,0.6135395367940267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,16384,1.8415662977430556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,16384,1.613827599419488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,12288,1.0844684176974826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,51200,5.3686951531304254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,65536,9.069533454047308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,51200,7.2315885755750875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,12288,1.30221070183648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,10240,1.1252507103814018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,10240,0.4003591007656521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,10240,1.113814247979058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,12288,0.45934131410386825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,10240,0.9833626217312283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,8192,0.8373137580023872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,8192,0.3090302149454753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,7168,0.6045920054117838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,12288,1.2247493531968858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,8192,0.7914889123704699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,7168,0.3375759919484456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,7168,0.7881466547648112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,6144,0.5168924331665039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,7168,0.6782471338907877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,6144,0.24034667015075684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,8192,0.8565457661946615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,6144,0.6557893223232693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,5120,0.20856266551547578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,6144,0.659805350833469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,5120,0.5487599902682835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,128,0.07623555262883504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,4096,0.17078755961524117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,4096,0.39960267808702254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,3584,0.38283289803398984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,3584,0.15814222229851618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,3584,0.35063910484313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,4096,0.33946045239766437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,3072,0.14214577939775255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,4096,0.43018044365776914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,3072,0.3272719913058811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,3072,0.30636265542772084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,5120,0.43479379018147785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,2560,0.12758755683898926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,2560,0.2773688899146186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,2048,0.17690043979220918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,2560,0.2588835557301839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,1536,0.148634672164917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,2048,0.11305510997772217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,2048,0.2266684373219808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,5120,0.48569154739379883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,2048,0.21393777264489067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,1024,0.09601155254575942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,1536,0.09767911169264053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,2560,0.2164106633928087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,768,0.06734577814737956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,1024,0.08373600244522095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,1024,0.12906489107343885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,1024,0.12595200538635254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,3072,0.2607911162906223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,512,0.05366311139530606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,768,0.10297511021296184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,768,0.10318133566114639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,256,0.035262223747041486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,3584,0.2745093239678277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,512,0.08005066712697347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,1536,0.17803111341264513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,128,0.030351112286249798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,1536,0.16841422186957467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,256,0.06008533636728922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,128,0.06521955463621351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,128,0.06131822533077664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,768,0.08167999982833862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,128,0.05054400033420987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,32,0.026368889543745253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,512,0.07271822293599446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,512,0.08283999893400404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,64,0.05948444207509359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,32,0.05866666634877523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,256,0.0682577755716112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,256,0.06639822324117024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,64,0.025789333714379206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,65536,2.2423973083496094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,51200,1.8079645368787978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,51200,3.4733929104275174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,16384,1.1026497946845162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,65536,4.855619642469618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,16384,0.5003226598103842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,16384,1.4147271050347223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,51200,4.320718977186415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,16384,1.3212693532307942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,51200,5.109023200141059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,12288,0.38045867284138996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,65536,5.602699703640408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,12288,0.7707609070671929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,12288,1.0671830707126193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,10240,0.331042660607232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,8192,0.497486220465766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,10240,0.8802168634202746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,65536,7.197469075520833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,8192,0.26631556616889107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,8192,0.7062320179409451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,7168,0.4078142113155789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,10240,0.7090249061584473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,8192,0.6453173425462511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,7168,0.23135911093817818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,6144,0.38125154707166886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,7168,0.6325902409023708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,6144,0.20459732744428846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,10240,0.8088249100579156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,5120,0.3593466546800401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,6144,0.5352480146620009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,6144,0.4816133181254069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,5120,0.19406044483184814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,5120,0.444818655649821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,5120,0.4065982235802545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,4096,0.1479520003000895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,4096,0.3633306556277805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,3584,0.24304178025987414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,7168,0.561590247684055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,4096,0.3349173333909776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,3584,0.15069866180419922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,3584,0.31268265512254506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,3584,0.29026489787631565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,4096,0.28916798697577584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,12288,0.9726817872789172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,3072,0.2696959972381592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,2048,0.1351919968922933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,2560,0.11059377590815227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,2560,0.2313777870602078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,2560,0.21437954902648926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,1536,0.10427644517686631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,2048,0.17625510692596436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,2560,0.16715910699632433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,3072,0.25383467144436306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,1536,0.13945421907636854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,1024,0.07358577516343859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,2048,0.1869893338945177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,1024,0.1061484416325887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,1024,0.10356177886327107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,768,0.05671822362475925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,1024,0.07514222462972005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,3072,0.12056533495585124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,768,0.07386222150590685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,768,0.08430755800671047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,768,0.08374133374955918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,1536,0.08640088637669881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,256,0.028615110450320776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,512,0.06552000178231134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,512,0.06553244590759277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,2048,0.09910933176676433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,256,0.05408089028464424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,256,0.06018310785293579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,256,0.04866133464707268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,1536,0.1465804444419013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,128,0.04828622274928623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,512,0.04197422332233853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,128,0.04168444540765551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,64,0.0461662213007609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,512,0.06720355484220716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,32,0.04497777753406101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,3072,0.1974551147884793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,128,0.024466666910383437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,128,0.05893511242336697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,64,0.021294222937689886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,32,0.02165688905451033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,65536,1.7600488662719727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,51200,1.3801946640014648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,51200,2.3954505920410156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,16384,0.7168622016906738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,16384,0.43395111295912003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,16384,1.114661322699653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,16384,1.0419378280639648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,51200,3.3641137017144094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,65536,3.229932361178928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,51200,3.6382675170898438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,65536,4.296162499321832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,65536,4.722759246826172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,12288,0.3122826682196723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,10240,0.42653687795003253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,12288,0.8203297721015083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,10240,0.2694755660163032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,8192,0.35681600040859646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,12288,0.764593760172526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,10240,0.6876053280300565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,8192,0.2077031135559082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,7168,0.29503199789259166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,8192,0.5739582379659017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,6144,0.24811466534932455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,7168,0.4906240039401584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,7168,0.44249510765075684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,6144,0.16698755158318412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,6144,0.42254045274522567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,6144,0.4138933287726508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,12288,0.5180835723876953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,5120,0.35210310088263613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,5120,0.13952533404032388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,7168,0.18395200040605333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,5120,0.3214035563998752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,3584,0.1558728880352444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,4096,0.11923378043704563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,4096,0.28245422575208873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,5120,0.2257200082143148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,4096,0.26092444525824654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,3072,0.12669155332777235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,3584,0.1095857752694024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,3584,0.24971999062432182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,3584,0.23139378759596083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,2560,0.11518222755855984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,3072,0.10079199737972683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,3072,0.21339555581410727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,8192,0.5081724590725368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,3072,0.2002462281121148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,2048,0.09635644488864475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,2560,0.09376977549658881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,2560,0.1709911160998874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,2048,0.08465422524346246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,2048,0.14782578415340847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,2048,0.1398115555445353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,10240,0.6436844401889378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,1024,0.05386133326424492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,1536,0.07563200261857775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,4096,0.17717688613467744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,1024,0.08273777696821424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,2560,0.1806346707873874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,768,0.04118666715092129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,1536,0.07076978021197848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,768,0.06560711065928142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,768,0.06521333588494195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,768,0.06570133235719469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,1536,0.1149102184507582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,1536,0.11040266354878743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,256,0.01984444426165687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,512,0.05140622125731575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,1024,0.08175466458002727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,256,0.040060444010628596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,256,0.04684888985421923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,128,0.016265778077973258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,256,0.03711377912097507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,64,0.01516355574131012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,128,0.04018311036957635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,32,0.015096889601813423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,128,0.03266222278277079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,64,0.03254310952292548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,512,0.05743288993835449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,32,0.031142221556769475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,512,0.025604445073339675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,128,0.03495999839570787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,512,0.051190220647388034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,1024,0.06698577933841281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,65536,1.684065818786621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,65536,2.468320846557617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,51200,1.8835644192165797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,16384,0.5925635761684841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,51200,1.3128809399074979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,16384,0.39831821123758954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,12288,0.3834346665276422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,65536,3.709905836317274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,16384,0.8989155027601453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,51200,3.043333265516493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,51200,2.8585376739501953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,10240,0.3230622344546848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,12288,0.3044915464189318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,12288,0.7223137749565972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,12288,0.6704266866048177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,10240,0.27868978182474774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,10240,0.5898897912767198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,65536,3.940118153889974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,7168,0.23067821396721733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,8192,0.47346221076117617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,10240,0.5592008696662055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,8192,0.4504568841722276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,6144,0.2059182193544176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,7168,0.413865778181288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,8192,0.28976623217264813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,7168,0.3924515512254503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,6144,0.3567208978864882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,5120,0.17708444595336914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,8192,0.20094754960801867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,6144,0.3374977906545003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,5120,0.13453777631123862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,5120,0.3017715613047282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,5120,0.283914672003852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,7168,0.17773155371348062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,4096,0.11672355069054498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,4096,0.24677512380811903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,3584,0.12442666954464382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,4096,0.23000534375508627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,3072,0.10946489042705959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,3584,0.10640355613496567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,6144,0.1588924460940891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,3584,0.2128195497724745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,3584,0.20263289080725777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,3072,0.09823466671837701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,3072,0.18478133943345812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,4096,0.14436977439456516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,3072,0.1777040031221178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,2560,0.09163822068108453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,2560,0.15679111745622423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,2560,0.14955022599962023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,1536,0.05478399991989136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,2048,0.12786489062839085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,2048,0.12452177206675212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,1024,0.03917155663172404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,1536,0.07376266850365533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,1536,0.09721599684821235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,2560,0.09659910864300197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,768,0.03035555614365472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,1024,0.07109066512849596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,2048,0.07568355401357015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,1024,0.07227022118038602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,768,0.06441689199871488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,768,0.05782755878236559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,16384,0.9750169118245443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,1536,0.09940889146592881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,512,0.04333422250217862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,512,0.04999377661281162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,512,0.04485866758558485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,1024,0.0662355555428399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,256,0.037775112522972956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,128,0.013526221944226159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,768,0.056243558724721275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,512,0.02015555567211575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,128,0.027792887555228338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,128,0.035655998521380954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,64,0.01240888900227017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,128,0.02831911047299703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,32,0.012799111505349478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,256,0.015311110350820752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,64,0.025748444928063288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,32,0.024597333537207708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,256,0.03292266527811686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,256,0.0317191117339664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,65536,1.7534888585408528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,51200,1.3901501761542425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,2048,0.0827582213613722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,65536,1.3324915568033855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,16384,0.4321137799157037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,16384,0.32721421453687877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,16384,0.8102515538533529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,51200,2.4993377261691623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,51200,1.0337262683444552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,51200,2.4062336815728083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,12288,0.2661324342091878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,65536,3.047397401597765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,10240,0.25016623073154026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,12288,0.2619617780049642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,12288,0.6076942019992405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,8192,0.2064159976111518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,10240,0.21761955155266655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,12288,0.5551902453104655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,10240,0.462066650390625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,7168,0.17177511586083305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,16384,0.7482977973090278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,8192,0.4080355432298448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,8192,0.3737848864661322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,6144,0.14609955416785345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,7168,0.14369689093695745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,7168,0.34865866767035586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,7168,0.32678577635023326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,5120,0.126383105913798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,6144,0.1261004474427965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,10240,0.5080622037251791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,6144,0.28124356269836426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,4096,0.10176533460617065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,5120,0.10855378044976129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,8192,0.1634897788365682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,5120,0.25342488288879395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,3584,0.09498488903045654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,5120,0.2356142203013102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,4096,0.09514577521218194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,4096,0.20898044109344482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,3072,0.08344444301393296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,4096,0.19165689415401885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,3584,0.08835644192165798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,3584,0.1799688869052463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,2560,0.06677333513895671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,3584,0.16948978106180826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,3072,0.08077155881457858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,3072,0.15902222527398005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,3072,0.1470951106813219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,2560,0.13230755594041613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,2560,0.1257502237955729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,2048,0.06924800078074138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,2048,0.1104115578863356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,2048,0.10282222429911296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,1536,0.08353955878151788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,1536,0.06330221891403198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,1024,0.027031110392676458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,6144,0.3019031153784858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,2560,0.0760071078936259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,1024,0.059803558720482715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,768,0.02145244512293074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,1024,0.056835558679368764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,1536,0.04239822096294827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,512,0.01555022266176012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,768,0.046796444389555186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,768,0.04788355694876777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,512,0.03557777735922072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,512,0.034712000025643244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,256,0.011735999749766456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,512,0.036532445086373225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,2048,0.056233776940239795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,128,0.010330666270520952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,256,0.030580444468392268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,1024,0.059374219841427274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,256,0.026076444321208533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,65536,3.167739656236437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,128,0.02146400014559428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,128,0.030184000730514526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,32,0.010282666318946415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,128,0.02423644396993849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,64,0.019357333580652874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,1536,0.08114222023222181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,256,0.026159110996458266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,65536,0.8736249075995551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,768,0.04957688848177592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,32,0.018232888645595975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,51200,0.7859519852532281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,64,0.010060444474220276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,65536,1.306014272901747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,16384,0.21161954932742646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,65536,2.5570977528889975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,65536,2.4746632046169705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,12288,0.16451199849446616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,51200,2.0146861606174045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,16384,0.6760800149705676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,16384,0.6223946677313911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,12288,0.2364648977915446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,51200,1.9214409722222223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,12288,0.5009111298455132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,12288,0.4660115771823459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,10240,0.20805954933166504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,10240,0.40505509906344944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,16384,0.30500978893703884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,10240,0.3866293430328369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,7168,0.09811466932296753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,8192,0.15405244297451443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,10240,0.1461520062552558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,8192,0.311938656700982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,6144,0.08558666706085205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,7168,0.2865217791663276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,8192,0.10120177931255764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,5120,0.06507200002670288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,6144,0.2465991179148356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,6144,0.23537866274515787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,51200,0.9928631252712674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,5120,0.21275199784172905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,4096,0.05940622091293335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,7168,0.13713955879211426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,7168,0.273305786980523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,4096,0.09170755412843491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,3584,0.053252445326911084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,4096,0.16005510754055446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,3584,0.08447111315197414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,3584,0.1470951106813219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,3072,0.046945777204301625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,5120,0.197272883521186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,3072,0.07843110958735149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,3072,0.12743288940853542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,2560,0.037096887826919556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,6144,0.12069600158267552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,2560,0.10825422075059678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,2048,0.029566221767001685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,8192,0.3268435531192356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,3584,0.14125066333346897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,2048,0.06686577532026503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,1536,0.02369244396686554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,3072,0.1226266622543335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,1536,0.06772355238596599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,2560,0.07303022013770209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,2560,0.10441155566109551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,5120,0.10541866885291205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,1536,0.06141599681642321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,1024,0.04775377776887682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,768,0.011777777638700275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,4096,0.17173333962758383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,768,0.03745244608985053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,768,0.03907910982767741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,1024,0.014846222268210517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,512,0.009690666364298927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,1536,0.06769066386752658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,512,0.027666666441493567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,256,0.007292444507280986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,2048,0.08591022094090779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,256,0.019277332557572257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,256,0.019863110449579027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,256,0.029862221744325426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,2048,0.08894666698243882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,128,0.015166223049163818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,1024,0.049157334698571094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,128,0.02953688965903388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,64,0.013136000268989138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,32,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,128,0.006960888703664144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,32,0.013912000589900546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,1024,0.05096266667048136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,64,0.007252444823582967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,65536,0.6916186544630262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,768,0.03903022077348497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,51200,0.5458658006456163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,512,0.029496000872717962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,512,0.03133422136306763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,16384,0.1797964440451728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,65536,1.3226017422146268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,65536,2.2798995971679688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,51200,0.9928035736083984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,16384,0.5772906409369575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,12288,0.13591910733117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,16384,0.30903734101189506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,12288,0.43343644671969944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,12288,0.23751023080613878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,51200,1.775879118177626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,10240,0.36349956194559735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,10240,0.11514933904012044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,10240,0.2108008861541748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,8192,0.09257066912121242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,8192,0.2953297826978895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,8192,0.15707022613949245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,7168,0.08274577723609076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,7168,0.2573360072241889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,16384,0.5530142254299587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,7168,0.13961599932776556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,6144,0.07161244418885973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,6144,0.22903111245897081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,12288,0.41784355375501847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,6144,0.1224000056584676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,5120,0.06143199735217624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,65536,2.180868360731337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,5120,0.1918133364783393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,5120,0.10602311293284099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,4096,0.050423112180497914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,5120,0.17983199490441215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,10240,0.35138577885097927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,4096,0.15538756052652994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,4096,0.14653600586785212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,51200,1.6991013420952692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,3584,0.13359822167290583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,6144,0.2137920061747233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,3072,0.038423111041386925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,3584,0.08517422278722127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,3584,0.0449813339445326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,4096,0.0918204453256395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,3072,0.11259110768636067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,2560,0.03276977936426798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,3072,0.07875377602047391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,8192,0.2846471203698052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,2048,0.027772444817754958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,2560,0.07266044616699219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,3584,0.12948444154527453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,2560,0.09541155232323541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,1536,0.018346667289733887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,2048,0.06585066848331027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,2048,0.07945511076185438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,2048,0.07779466443591647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,1024,0.009687999884287516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,3072,0.11857955985599095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,1536,0.06020000245836046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,1536,0.05962044662899441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,1024,0.043440000878440015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,1024,0.04229155513975355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,1024,0.04262666569815742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,2560,0.09773600101470947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,512,0.006960888703664144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,768,0.0321777794096205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,768,0.033723556333118014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,512,0.02365422248840332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,256,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,768,0.008393777741326226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,256,0.015836444165971544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,256,0.029887109994888306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,768,0.03432000014517043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,256,0.017260443833139207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,7168,0.2481048901875814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,512,0.03126310971048143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,128,0.029669334491093952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,512,0.025536000728607178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,32,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,64,0.010016000105275048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,32,0.01037688884470198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,128,0.004991111242108875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,65536,0.649238215552436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,128,0.011737777955002256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,1536,0.06075644493103027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,51200,0.497411568959554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,12288,0.12396621704101562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,16384,0.5632150967915853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,12288,0.4297164546118842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,10240,0.10339733627107407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,64,0.004993777722120285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,8192,0.08583111233181423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,7168,0.0740977790620592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,16384,0.16096356179979113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,6144,0.06421511040793525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,10240,0.3568728764851888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,51200,1.7304977840847438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,8192,0.2974284489949544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,4096,0.045349333021375865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,7168,0.2534124321407742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,3584,0.0397093329164717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,5120,0.1824373271730211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,3072,0.034154666794670954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,4096,0.15222222275204128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,2560,0.029142220815022785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,3584,0.13064889113108316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,2048,0.023919999599456787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,3072,0.1128657791349623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,1536,0.013078221844302284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,2560,0.09705777963002522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,1024,0.008672888908121321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,2048,0.07628533575269911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,6144,0.2246097723642985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,1536,0.057428446080949574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,768,0.007648000286685095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,512,0.0063164445261160536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,1024,0.03951111104753282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,256,0.004950222041871813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,768,0.03090488910675049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,128,0.004554666578769684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,65536,2.218620512220595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,256,0.014270222849316068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,5120,0.05509066581726074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,64,0.004222222086456087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,32,0.0047644442982143825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,128,0.010324444207880232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,64,0.009352888498041365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,32,0.009877333210574256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,512,0.022096888886557683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,65536,0.7050106790330676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,16384,0.16829955577850342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,12288,0.12882044580247667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,16384,0.5645244386461046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,10240,0.10893955495622422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,12288,0.43736889627244735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,8192,0.09072622325685288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,51200,1.7243199878268773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,7168,0.07787377966774835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,65536,2.22377692328559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,51200,0.5655768712361654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,6144,0.06713866525226168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,5120,0.0562266641192966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,8192,0.2969662348429362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,7168,0.25325245327419704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,6144,0.21813066800435385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,128,0.019578667150603402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,3072,0.03623200125164456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,3584,0.13229155540466309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,2560,0.030555556217829388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,10240,0.35619378089904785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,3072,0.11425422297583686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,4096,0.04599822229809231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,1536,0.01313688854376475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,2560,0.09617688920762803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,1024,0.007674666742483775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,2048,0.07875555753707886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,768,0.006579555571079254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,4096,0.15215288268195257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,1536,0.05961155229144626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,512,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,1024,0.04192444350984362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,768,0.032630221711264715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,128,0.003536000019974179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,2048,0.025590222742822435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,64,0.0035537779331207275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,128,0.010383110907342698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,3584,0.04048533240954081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,64,0.00959022260374493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,256,0.004568888909286923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,512,0.02307555576165517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,256,0.014512888259357877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,32,0.0032968889507982465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,32,0.00982844415638182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,5120,0.18758311536577013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,128,0.018975110517607797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,16384,22.75089179144965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,16384,8.898661295572916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,51200,26.172403971354168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,16384,29.660698784722225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,12288,6.558110978868272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,12288,17.97126939561632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,16384,42.048485649956596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,12288,21.950581868489582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,12288,30.863969590928818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,10240,5.326524522569445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,10240,16.50055609809028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,10240,18.889346652560764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,8192,10.545973883734808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,10240,25.53360663519965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,51200,68.18707953559029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,8192,4.2045512729220915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,7168,3.681085374620226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,8192,14.805494520399305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,7168,9.073716057671442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,8192,21.822162204318577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,6144,8.482026841905382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,7168,12.63365258110894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,6144,3.1173057556152344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,7168,17.537423027886284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,6144,12.410281711154513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,6144,10.770237392849394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,5120,2.595759073893229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,5120,6.870424906412761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,5120,7.829515245225694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,4096,2.022094302707248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,4096,5.454043494330512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,5120,7.7617094251844625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,4096,5.726582421196832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,3584,1.8083226945665147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,3584,4.477507697211371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,4096,5.626082526312934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,3584,5.060305701361762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,3584,4.781399197048612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,3072,3.9802364773220487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,3072,1.6058052910698786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,3072,4.247178819444445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,3072,4.15486568874783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,2560,1.3359066645304363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,2560,3.8288781907823353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,2560,3.4120089213053384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,2048,3.1440177493625217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,2560,3.4638036092122397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,2048,2.8471031188964844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,1536,2.041486316257053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,2048,2.784009721544054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,2048,1.192795541551378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,1536,2.1473990546332464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,1536,2.273620393541124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,1024,1.5410417980617945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,1024,0.6512115796407064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,1024,1.4556951522827148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,768,1.210311147901747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,1536,0.821964422861735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,768,1.218113793267144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,768,0.547095086839464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,1024,1.573280864291721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,768,1.2373502519395616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,512,0.8550737698872884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,512,0.47797245449490017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,512,0.9878809187147352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,512,0.9158230887518989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,256,0.4329395559098985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,256,0.8996915817260742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,256,0.7500960032145182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,256,0.5710719956292046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,128,0.4155013296339247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,64,0.3529191017150879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,128,0.5455679893493652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,128,0.4227377838558621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,64,0.8678453233506945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,32,0.9786560270521375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,128,0.8403884569803873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,32,0.3610195583767361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,51200,127.7190890842014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,65536,27.58743964301215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,51200,95.95753309461806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,51200,21.35293748643663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,51200,52.324608696831596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,16384,16.21118333604601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,65536,71.06075032552083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,16384,6.938138749864366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,16384,22.812795003255207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,16384,33.09638807508681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,12288,14.272589789496529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,12288,5.039723714192708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,12288,17.14753892686632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,51200,73.78073459201389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,10240,10.65517086452908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,51200,99.80900065104167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,12288,24.089548746744793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,10240,4.178541395399305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,8192,3.4112639957004123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,8192,10.322866651746962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,10240,14.581853230794271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,10240,19.91770765516493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,8192,11.880601671006945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,7168,8.359574211968317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,7168,2.938078138563368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,8192,17.043672349717884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,6144,2.4580205281575522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,6144,5.924430423312717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,7168,9.78102789984809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,7168,13.721500820583769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,6144,8.517588297526041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,5120,5.491143120659722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,5120,2.0192559560139975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,5120,6.300791846381293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,6144,9.57883029513889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,4096,1.55657778845893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,5120,5.867191314697266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,4096,4.330367194281684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,4096,4.651343875461154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,4096,4.39064449734158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,3584,1.3300915824042427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,3584,3.9730300903320312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,3584,3.9309476216634116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,3584,3.749808841281467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,3072,1.239000850253635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,3072,3.2247528500027127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,65536,133.54306369357639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,3072,3.2526639302571616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,2560,2.7135876549614797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,2560,0.9788729349772135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,3072,3.2267110612657337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,2560,2.71380975511339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,2048,0.9059279759724935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,2048,2.45871819390191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,2048,2.2345973120795355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,1536,1.8836665683322484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,1536,0.658532460530599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,2560,2.759138743082682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,1024,1.1344711515638564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,1536,1.6893431345621746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,2048,2.1608844333224826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,1024,0.6503049002753364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,1024,1.2139173083835177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,768,0.8637706438700358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,768,0.435007996029324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,768,0.9250684314303927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,768,0.9243582619561089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,512,0.6919706662495931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,512,0.7752426465352377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,512,0.3673715591430664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,256,0.5566435390048557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,1536,1.6291173299153645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,1024,1.2171724107530382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,256,0.6975244416130914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,256,0.5706186824374729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,64,0.27041421996222603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,128,0.6554995642768012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,128,0.42887555228339297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,32,0.2815715471903483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,128,0.32412976688808864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,256,0.4526195526123047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,64,0.6779680252075195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,128,0.3286951118045383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,512,0.7268275684780545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,32,0.7667333285013834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,65536,94.20405409071181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,65536,10.689268324110243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,16384,5.5335939195421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,51200,16.71401638454861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,51200,8.20689222547743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,65536,20.576800876193577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,16384,2.5951048533121743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,12288,4.48689227634006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,16384,7.549604627821181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,51200,23.40731472439236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,12288,1.8007734086778429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,16384,10.716124640570747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,12288,7.8231396145290795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,10240,3.720374213324653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,12288,5.430985768636067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,10240,1.5793644587198894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,65536,31.073137071397568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,10240,4.932592180040148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,8192,3.3074525197347007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,10240,6.489882575141059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,51200,32.403479682074654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,8192,0.9566124810112847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,7168,0.8264693154229058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,7168,2.610992007785373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,8192,5.5296630859375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,8192,3.6543655395507812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,6144,0.6704950862460666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,7168,3.2106844584147134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,7168,4.4344461229112415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,5120,1.8691147698296442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,6144,2.1714950137668185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,6144,3.1379138098822703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,5120,0.5767911275227865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,5120,1.9524887932671442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,5120,1.938042746649848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,6144,2.7886488172743054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,4096,1.431653340657552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,4096,1.3733332951863606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,4096,1.4150578180948894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,3584,0.5102044211493598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,3584,1.3603813383314345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,3584,1.1985670725504558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,3584,1.189652442932129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,3072,0.4654080073038737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,3072,0.9928853776719836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,3072,0.9462098015679253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,4096,0.533223999871148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,2560,0.31715912289089626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,3072,1.1045413547092013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,2560,0.7750230895148383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,2048,0.27860087818569607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,1536,0.49155023362901473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,2048,0.631131543053521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,2560,0.841318236456977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,1536,0.21368000242445204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,65536,44.512529161241325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,1024,0.3674320114983453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,1536,0.5123768912421333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,1024,0.2000782224867079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,1024,0.36923199229770237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,1536,0.5519110891554091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,2048,0.6791769133673774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,768,0.15211999416351318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,768,0.30032266510857475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,2560,0.8480275472005209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,512,0.25652623176574707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,512,0.1303573316997952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,512,0.24145955509609648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,256,0.1551520029703776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,768,0.3165377775828044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,256,0.11833600203196208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,512,0.19592532846662733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,256,0.18949245082007515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,1024,0.39423378308614093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,128,0.2151262230343289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,128,0.11944266160329182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,64,0.0887226661046346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,32,0.09329244163301255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,2048,0.6702382299635145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,256,0.22982578807406953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,768,0.26192087597317165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,128,0.14562133948008218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,128,0.10733599795235528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,64,0.22136267026265463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,32,0.24975734286838105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,65536,8.10229746500651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,51200,6.367198096381293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,51200,12.077414618598091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,16384,4.070346832275391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,65536,15.548439873589409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,16384,1.9528950585259333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,16384,8.117471483018663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,12288,3.281203587849935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,16384,5.825808207194011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,12288,1.0953857633802626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,51200,18.064224243164062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,12288,5.86477788289388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,12288,4.259337955050998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,10240,2.981489817301432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,10240,0.9275200102064344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,65536,22.4889899359809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,10240,3.551607131958008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,8192,2.358220418294271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,8192,0.7762319776746961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,10240,4.880425771077474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,8192,2.962195502387153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,51200,24.350521511501736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,8192,3.0846701727973094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,7168,0.5949333508809408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,7168,2.066505855984158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,6144,0.5297199885050455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,6144,1.6205146577623155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,7168,2.3385609520806208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,5120,1.3857911427815754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,6144,1.9858133527967665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,5120,0.5380115509033203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,5120,1.3112160364786785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,7168,2.3166845109727645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,6144,1.617637316385905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,4096,0.35940888192918563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,5120,1.4808852937486436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,3584,0.9242791069878472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,4096,0.9989190631442599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,3584,0.41194134288363987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,4096,1.016783078511556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,3584,0.8648106786939832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,3584,0.8500035603841146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,4096,1.0585270987616646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,3072,0.8114408916897244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,3072,0.749784893459744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,2560,0.624368879530165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,3072,0.7382453282674154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,2560,0.2806497679816352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,2048,0.46472443474663633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,2048,0.20559377140469023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,3072,0.28630044725206166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,1536,0.3669555452134874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,2048,0.4971546596950955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,1536,0.4057164457109239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,1024,0.26551821496751576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,1536,0.3739635679456923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,2560,0.6532204416063097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,2560,0.611396418677436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,1024,0.297818660736084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,768,0.22930577066209581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,2048,0.5540382067362467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,768,0.1177448829015096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,512,0.15928088294135198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,1536,0.2016622225443522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,768,0.22899288601345488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,65536,33.58120388454861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,256,0.11323822869194879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,512,0.10433333449893528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,512,0.19323821862538657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,512,0.18456621964772543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,128,0.08453866508271958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,256,0.09587466716766357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,256,0.17131555080413818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,256,0.14486755265129939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,128,0.09631199969185723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,128,0.1623928944269816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,768,0.23531201150682238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,64,0.16243288252088758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,32,0.1776862276924981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,1024,0.12989866733551025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,64,0.07274222373962402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,32,0.07188444667392306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,1024,0.3877386781904433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,128,0.11232088671790229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,65536,6.714827643500434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,51200,4.84080547756619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,16384,3.4711901346842446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,51200,10.859632703993055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,65536,14.331854926215279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,16384,6.874568091498481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,16384,1.4536657333374023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,16384,4.602416144476996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,12288,0.9017040464613172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,12288,3.0034569634331594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,51200,15.250325520833334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,12288,3.501514646742079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,12288,4.873685201009114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,10240,0.7190373208787707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,10240,2.459546618991428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,65536,20.07302517361111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,10240,2.957651562160916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,51200,20.501634385850696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,8192,0.5722453329298232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,8192,2.223564359876845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,10240,3.4369644588894315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,8192,1.9782488081190321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,7168,1.594198226928711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,7168,0.5444729063245985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,8192,2.0798382229275174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,7168,1.6035244199964735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,6144,0.4272622267405192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,6144,1.4510587056477864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,7168,2.1107218000623917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,6144,1.2645866605970595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,5120,1.1314853032430012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,5120,0.4463013543023004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,6144,1.608003510369195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,4096,0.31345423062642414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,4096,1.0065039528740776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,4096,0.8205306794908312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,3584,0.7984426816304525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,5120,1.0461031595865886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,3584,0.2804959879981147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,3584,0.7441697650485568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,5120,1.185256004333496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,3584,0.7028142081366645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,3072,0.2466053432888455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,3072,0.6346853574117025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,3072,0.6182444360521104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,2560,0.5303342077467177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,2560,0.2109395530488756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,2560,0.547220442030165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,2048,0.422776010301378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,2560,0.5079769028557671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,2048,0.17735200458102754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,3072,0.6839911142985026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,1536,0.3304995430840386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,2048,0.40370933214823407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,1536,0.14397422472635904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,1024,0.2308879958258735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,4096,0.8865280151367188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,1024,0.13461421595679388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,1024,0.25641600290934247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,768,0.18425867292616102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,1024,0.23772088686625162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,512,0.12836266888512507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,2048,0.5045724444919163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,65536,28.21383327907986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,1536,0.34201778305901426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,512,0.16387911637624106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,256,0.09529333644443089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,512,0.15663022465176052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,256,0.08304799927605523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,256,0.14488800366719565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,128,0.07008710834715101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,768,0.10116533438364665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,768,0.20859377914004854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,768,0.19415822294023302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,128,0.08400355445014106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,128,0.0956160028775533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,512,0.09210044145584106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,64,0.13663200537363687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,32,0.14464888307783338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,256,0.12295111020406087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,64,0.06477955314848158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,32,0.06770488950941297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,1536,0.31609421306186253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,128,0.13679111003875732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,65536,5.220047844780816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,51200,4.182620578342013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,16384,2.7376310560438366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,65536,10.35322740342882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,51200,9.707329644097223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,16384,0.9966275956895617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,51200,10.124901665581598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,16384,3.112385855780708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,12288,2.0577680799696183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,16384,4.799093458387587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,12288,0.6805066532558866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,65536,12.692599826388888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,12288,2.3181964026557074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,10240,0.567072868347168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,10240,1.9407261742485895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,12288,2.9564772711859812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,10240,1.9551565382215712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,8192,0.44626135296291775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,10240,1.8620755937364366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,8192,1.4418346616956923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,7168,1.240495999654134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,8192,1.3981777826944988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,8192,1.3445555369059246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,7168,0.4265448782179091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,7168,1.250240855746799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,7168,1.1573777728610568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,6144,1.1406390931871202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,6144,0.34502310223049587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,51200,15.389424641927084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,6144,1.031002680460612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,6144,0.9647777345445422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,5120,0.29747022522820366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,4096,0.6918204625447592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,5120,0.8032702340020074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,5120,0.9210426542494031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,4096,0.30230667856004506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,4096,0.7060977617899576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,3584,0.5295262336730957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,4096,0.6148239771525065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,5120,0.8283368746439616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,3584,0.23011112213134766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,3584,0.5912151336669922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,3072,0.541885322994656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,2560,0.4122142261928982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,3072,0.5297937923007542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,3072,0.4728710916307237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,2560,0.20245422257317436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,2048,0.33234045240614146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,2560,0.5385804706149632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,2048,0.1631315549214681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,2048,0.36290756861368817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,1536,0.25801155302259654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,2048,0.33939822514851886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,1536,0.13296622700161403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,1536,0.28817688094245064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,1536,0.26164976755777997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,1024,0.20612533887227377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,1024,0.1987688938776652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,3584,0.5368213123745388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,65536,22.60940382215712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,768,0.1630906661351522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,2560,0.4303022225697835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,768,0.1618408891889784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,512,0.10052977667914496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,512,0.07928266790178087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,512,0.13433867030673557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,256,0.07194133599599202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,768,0.14548267258538142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,256,0.07205422057045831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,1024,0.18533599376678467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,128,0.11097777552074856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,512,0.13004533449808756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,3072,0.20163467195298937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,64,0.04704888992839389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,128,0.07196444272994995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,128,0.08072888851165771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,32,0.04874933428234524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,128,0.05677600039376152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,64,0.11110577980677287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,32,0.11486132939656575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,1024,0.09453244341744317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,256,0.11797066529591878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,768,0.08797599871953328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,256,0.10153422090742324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,65536,4.480898539225261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,51200,3.5771653917100696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,16384,2.5652614169650607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,51200,8.416450500488281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,65536,9.558519999186197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,16384,0.8173039754231771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,16384,3.8726374308268228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,51200,8.750186496310764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,16384,2.668970743815104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,12288,1.9674701690673828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,12288,0.5910293261210123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,65536,11.181078592936197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,12288,2.26712883843316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,12288,1.9724542829725478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,10240,1.8620925479465062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,10240,0.5089777840508355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,10240,1.5536613464355469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,51200,12.543754577636719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,8192,0.403547551896837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,8192,1.2471929126315646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,10240,1.6237271626790364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,8192,1.2256684833102756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,7168,1.1545030805799696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,7168,1.0426461961534288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,8192,1.3224728902180989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,7168,1.008676422966851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,6144,0.7786177529229058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,6144,0.3186711205376519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,6144,0.9197742674085828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,5120,0.27020266320970326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,6144,0.8175226847330729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,7168,0.3548399872250027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,5120,0.7608426411946615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,5120,0.6987102296617297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,4096,0.2224737803141276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,4096,0.6009075376722548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,3584,0.5059982405768501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,5120,0.7404889000786675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,3584,0.5258293151855469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,3584,0.24083111021253797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,3584,0.47851912180582684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,4096,0.606337759229872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,3072,0.1781768931282891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,3072,0.45534223980373806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,2560,0.3938799964057074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,4096,0.5483946800231934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,3072,0.42836888631184894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,2560,0.1602133380042182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,2048,0.3206720087263319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,2560,0.35289687580532497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,2048,0.317267550362481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,1536,0.22635910246107313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,2048,0.3012773460812039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,1536,0.24944978290134004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,1536,0.11019466982947455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,65536,19.807566324869793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,1024,0.15302844842274985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,2560,0.39073954688178164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,1536,0.24337866571214464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,768,0.1125288936826918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,1024,0.09546755419837104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,1024,0.1825528939565023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,3072,0.4515013164944119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,768,0.07977333333757189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,768,0.14502310752868652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,768,0.14385333326127794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,512,0.07208977805243598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,512,0.11701422267489964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,256,0.1038657757971022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,128,0.05490577883190579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,512,0.09055022398630778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,128,0.09797777732213338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,256,0.06697599755393134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,512,0.1191457774904039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,64,0.04658044377962748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,128,0.07212977939181857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,32,0.048469334840774536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,64,0.09796177678638035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,32,0.10143199894163345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,2048,0.13173333803812662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,128,0.06616533464855619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,256,0.0902151134279039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,1024,0.17880533801184761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,256,0.06633689006169637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,65536,3.956984202067057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,51200,2.901704788208008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,16384,2.1624062856038413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,51200,6.507150014241536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,65536,8.564817640516493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,16384,0.7046826680501302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,16384,3.0546817779541016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,16384,2.2527857886420355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,12288,1.5848461786905925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,12288,0.5101564195421007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,65536,9.904106140136719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,12288,1.7393457624647353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,51200,7.5866750081380205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,12288,1.623652458190918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,10240,1.501532448662652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,8192,1.0027164883083768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,10240,1.2803644604153104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,8192,1.0370542738172743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,10240,0.4351377752092149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,8192,0.4266506565941705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,8192,1.0018142064412434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,7168,0.8719040022956001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,7168,0.3287022113800049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,10240,1.3470142152574327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,6144,0.7158453199598523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,7168,0.8962266710069445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,6144,0.27327556080288357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,51200,11.747772216796875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,6144,0.7943093511793348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,5120,0.7290657891167535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,5120,0.23970399962531197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,5120,0.5875786675347222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,4096,0.5159644550747341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,7168,0.8396959834628634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,4096,0.19235377841525605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,4096,0.5455430878533257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,4096,0.48076799180772567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,3584,0.542834652794732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,65536,15.943147447374132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,3584,0.17577955457899305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,3584,0.4603413475884332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,3072,0.1560693316989475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,2560,0.329711119333903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,3072,0.5233208868238661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,3072,0.372017780939738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,2560,0.1372764507929484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,2048,0.27236888143751353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,2560,0.30870310465494794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,3072,0.3551359971364339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,5120,0.6461688677469889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,2048,0.2765555646684435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,2048,0.2598257859547933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,1536,0.21790577305687797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,2560,0.3335582150353326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,1536,0.2039937840567695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,1024,0.13675822152031794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,3584,0.41751734415690106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,1536,0.21284711360931396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,768,0.11441955301496719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,6144,0.7386159896850586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,1536,0.09599466456307305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,2048,0.11643111705780029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,1024,0.15996800528632268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,768,0.12694756189982095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,512,0.07335378064049615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,1024,0.15594844023386636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,256,0.057337780793507896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,512,0.06593155860900879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,512,0.10451289017995198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,512,0.10177066591050889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,768,0.12707911597357854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,256,0.06004089117050171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,256,0.07867466741138034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,768,0.0731182230843438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,128,0.08509955803553264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,128,0.06385599904590182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,1024,0.07816000117195977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,64,0.08467555708355373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,32,0.0862426691585117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,128,0.04368799924850464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,256,0.09054044220182632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,64,0.037567999627855085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,32,0.03899466660287645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,128,0.05985866652594673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,65536,3.065817726982964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,51200,2.664552900526259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,16384,1.7453182008531358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,51200,5.5346772935655375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,16384,0.6005982293023003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,16384,2.240217844645182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,51200,6.239251454671224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,16384,1.8832168579101562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,12288,1.3038373523288302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,65536,8.852933247884115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,12288,0.4519786834716797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,10240,1.1294630898369682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,51200,9.092788696289062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,10240,1.2420097986857097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,10240,0.44031288888719344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,10240,1.076883527967665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,8192,0.9126870897081164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,8192,0.3086826536390516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,8192,0.8918595843844943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,8192,0.8255075348748101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,7168,0.8364924324883355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,7168,0.2688853210873074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,7168,0.7894355456034342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,12288,1.3731955422295465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,7168,0.6990968916151258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,6144,0.663807127210829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,65536,8.278650919596354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,6144,0.6633671124776205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,6144,0.6107430987887913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,5120,0.20370754930708143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,5120,0.6412533124287924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,4096,0.43875911500718856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,5120,0.4983990987141927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,4096,0.19051378303103975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,4096,0.4417288833194309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,5120,0.5222800042894151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,4096,0.41103288862440324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,12288,1.3312293158637154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,3584,0.1520577801598443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,3072,0.32602132691277397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,3584,0.35781510670979816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,6144,0.23503110143873426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,3072,0.14163466294606528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,2560,0.2771057817671034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,3584,0.3916035493214925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,2560,0.28805245293511283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,2048,0.23148534033033583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,65536,12.28717041015625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,2560,0.2687884436713325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,2048,0.10208800103929307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,2048,0.22590488857693145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,3072,0.338658677207099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,1536,0.1867964400185479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,3072,0.3126675552792019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,1024,0.11657778422037761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,1536,0.17546844482421875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,1024,0.07038578059938219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,1024,0.13746933142344156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,768,0.08845688899358113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,1024,0.13551910718282065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,768,0.06584888696670532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,1536,0.08470755815505981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,768,0.1090444458855523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,3584,0.4053155581156413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,512,0.08931199709574382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,512,0.05914577510621813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,1536,0.17334844006432426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,256,0.07696266969045003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,768,0.10906844668918186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,256,0.06744711266623603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,512,0.06401955419116549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,128,0.03728355632887946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,256,0.05406399899058872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,128,0.05402222275733948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,128,0.07207110855314466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,64,0.0355697770913442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,32,0.03738488753636678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,64,0.07142044438256158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,32,0.07241688834296332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,2560,0.12028088834550645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,128,0.055267555846108325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,2048,0.23640534612867567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,256,0.04793955551253425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,65536,2.403353797064887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,512,0.08755911058849758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,65536,5.531958262125651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,51200,1.9137698279486761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,16384,1.379683600531684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,51200,4.652709537082248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,16384,0.5206809043884277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,65536,6.518458472357856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,16384,1.578394677903917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,16384,1.5048408508300781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,12288,0.368430217107137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,12288,1.1614071528116863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,51200,4.864125145806207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,12288,1.0380498038397896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,51200,6.858935038248698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,10240,0.9199857711791992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,8192,0.756023089090983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,10240,0.8789031240675184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,8192,0.7416008843315972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,8192,0.33401955498589414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,8192,0.6630293528238932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,10240,0.33974221017625594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,65536,11.20594448513455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,7168,0.22454934649997285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,7168,0.6495395766364204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,6144,0.4842560026380751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,7168,0.5863564279344347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,6144,0.19747200277116564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,6144,0.5680826505025228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,5120,0.4327093230353461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,6144,0.4912426736619737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,5120,0.1702951060401069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,5120,0.4577600161234538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,4096,0.34168622228834367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,5120,0.4123155540890164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,4096,0.14296000533633763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,3584,0.284097777472602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,10240,0.9006399578518338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,4096,0.33776354789733887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,3584,0.3255724377102322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,3072,0.2796871132320828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,12288,1.1068301730685763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,3584,0.2976248794131809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,3072,0.12673955493503147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,2560,0.20588089360131157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,3072,0.25938932100931805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,2560,0.10399911138746475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,4096,0.3686408996582031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,2560,0.23899732695685494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,2560,0.2228391170501709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,3584,0.14337066809336343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,1536,0.1432044506072998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,2048,0.19852888584136963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,2048,0.18684177928500703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,1536,0.07379911343256633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,1536,0.15532533327738443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,1536,0.14734933111402723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,2048,0.1758426692750719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,1024,0.11422933472527398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,768,0.07524444659550984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,2048,0.0881155530611674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,768,0.09083644549051921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,768,0.05827111005783081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,512,0.056826664341820605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,512,0.07431644201278687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,1024,0.0616204473707411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,256,0.03563822309176127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,512,0.05282666616969638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,7168,0.5503875414530436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,256,0.04792266752984789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,768,0.09121155738830566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,128,0.05894222524431017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,512,0.07297066847483318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,1024,0.11139822006225586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,64,0.025528000460730657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,128,0.04695111182000902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,32,0.026156443688604567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,256,0.056136886278788246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,32,0.05854755640029907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,128,0.04792266752984789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,256,0.06367555591795179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,128,0.030603554513719346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,1024,0.08911022212770249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,65536,2.09386232164171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,3072,0.28408977720472545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,65536,5.056398179796007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,64,0.05811466773351034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,51200,4.160165150960286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,16384,1.2555173238118489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,65536,5.763665941026475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,51200,1.8242496914333766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,16384,1.3542088402642145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,16384,0.4395306640201145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,12288,0.9163946575588651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,65536,7.945814344618056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,12288,0.3403724564446344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,16384,1.325320879618327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,51200,5.934571584065755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,51200,4.386097802056207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,12288,0.8919217851426867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,10240,0.29525687959459096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,10240,0.8386649025811089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,8192,0.663541316986084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,10240,0.7721698019239637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,8192,0.6719457838270398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,8192,0.597175121307373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,7168,0.633985784318712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,12288,1.0288462109035916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,7168,0.2092426617940267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,6144,0.4518204265170627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,7168,0.5190889040629069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,6144,0.18152799871232775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,8192,0.23200088077121309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,6144,0.5610933303833008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,6144,0.44459909862942165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,5120,0.4121377733018663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,5120,0.1579937802420722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,4096,0.33420621024237734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,5120,0.3829653263092041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,4096,0.13162310918172201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,4096,0.3349422348870172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,7168,0.6970524258083768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,3584,0.11751022603776719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,10240,0.8372693591647677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,3584,0.27103998925950795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,5120,0.418870210647583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,3072,0.23555911911858451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,2560,0.21246043841044107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,4096,0.34714044464959043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,3584,0.2928248776329888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,2560,0.093941330909729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,2560,0.20942933029598662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,2048,0.17211466365390352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,3072,0.10713422298431396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,2048,0.08131911357243855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,2048,0.17733777893914116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,1536,0.1248062186770969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,2048,0.1681466632419162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,1024,0.08833422263463338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,1536,0.13995644781324598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,1536,0.13350400659773085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,2560,0.21551911036173502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,768,0.06323289208941989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,1024,0.0579377810160319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,1024,0.10099822282791138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,3072,0.25269688500298393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,512,0.04782755507363213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,768,0.0818835563129849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,3584,0.2721173233456082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,1536,0.06821244292789035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,512,0.04890666736496819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,3072,0.22713867823282877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,1024,0.102801779905955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,256,0.056699554125467934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,128,0.029462221595976088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,256,0.04523555437723795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,768,0.05496622125307719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,768,0.08298844761318631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,128,0.0525875555144416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,128,0.0448053346739875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,128,0.04245333207978142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,512,0.06667911344104342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,512,0.06639911068810357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,64,0.05154488815201653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,32,0.05201955636342367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,256,0.051017777787314526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,64,0.02549955579969618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,256,0.03612000081274245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,65536,1.802809821234809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,32,0.02677244444688161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,16384,1.0359644360012479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,51200,3.438621309068468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,51200,1.3850729200575087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,65536,4.554317474365234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,16384,0.4685795572068956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,16384,1.1899208492702908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,51200,3.747346666124132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,12288,0.8201280170016818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,65536,4.844156477186415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,16384,1.132449785868327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,12288,0.303276432885064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,12288,0.8955396016438802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,12288,0.8145075374179416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,10240,0.7440195613437228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,10240,0.6539537641737196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,8192,0.5374168819851345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,65536,7.013281928168403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,51200,6.158572302924262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,8192,0.21996266312069365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,8192,0.5957519743177626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,10240,0.6446462207370335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,8192,0.5288409127129449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,10240,0.2607048882378472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,6144,0.36562310324774844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,7168,0.7216906547546387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,7168,0.46491644117567277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,6144,0.16034666697184244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,6144,0.44709157943725586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,5120,0.3206542332967122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,6144,0.39487822850545246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,7168,0.45613686243693036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,5120,0.37067556381225586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,7168,0.18759911590152314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,5120,0.33356356620788574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,4096,0.2893137666914198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,4096,0.11579022142622207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,3584,0.22459022204081217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,4096,0.2681484487321642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,3072,0.1907973289489746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,3584,0.2580924563937717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,3584,0.2382000022464328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,3072,0.0948257777425978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,2560,0.1716151105033027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,3072,0.21006755034128824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,2560,0.08575999736785889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,2560,0.18988800048828125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,2048,0.13246222337086996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,2560,0.18016710546281603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,3584,0.10496888558069865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,1536,0.10798222488827175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,2048,0.15706400076548258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,2048,0.14916355080074734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,5120,0.1424728896882799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,1536,0.12451199690500896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,1536,0.11784356170230442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,1024,0.09130399756961399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,1024,0.05345244540108574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,1024,0.08979911274380153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,2048,0.07556621895896064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,768,0.07263555791642931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,4096,0.25363821453518337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,1536,0.06294222010506524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,512,0.042661332421832614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,768,0.05048799845907423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,256,0.02921244502067566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,512,0.04541866646872627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,512,0.05897066990534464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,256,0.049807998869154185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,256,0.04177688890033298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,256,0.04479111234347025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,3072,0.22471645143296984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,1024,0.06953155332141452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,128,0.046329776446024575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,768,0.07383822070227729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,128,0.04177066683769226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,128,0.03795466820398966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,512,0.05928977992799547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,768,0.057842665248446994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,32,0.04483644498719109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,128,0.023823999696307715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,64,0.02088088956144121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,32,0.02145333256986406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,64,0.04485599862204658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,65536,1.553864902920193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,51200,1.1871342129177518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,16384,0.9669831593831381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,51200,3.0024933285183377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,65536,3.769112057156033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,16384,0.3793182108137343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,12288,0.6468044386969672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,65536,4.212835523817274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,16384,0.956284417046441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,12288,0.2722497781117757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,51200,4.36625501844618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,12288,0.7039644453260633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,16384,1.125292460123698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,65536,6.009944068060981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,51200,3.2714390224880643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,10240,0.23515110545688203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,10240,0.6346053547329372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,8192,0.45180712805853945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,10240,0.5733662181430393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,7168,0.3714382118648953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,10240,0.5450684229532877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,8192,0.5190844535827637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,8192,0.4564835760328505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,7168,0.16715378231472441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,6144,0.3300702306959364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,7168,0.4981848928663466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,7168,0.4046035607655843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,6144,0.1446897718641493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,5120,0.2812817891438802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,6144,0.38463910420735675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,6144,0.34673155678643125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,8192,0.18239733907911512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,4096,0.23858133951822916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,5120,0.2924079895019531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,4096,0.2555404504140218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,3584,0.1867226627137926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,3584,0.2257902092403836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,3584,0.211517333984375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,3584,0.09414222505357529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,3072,0.1580924457973904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,4096,0.12711289193895128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,4096,0.23786844147576225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,3072,0.08568888902664185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,2560,0.130803558561537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,5120,0.1258631149927775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,5120,0.32472888628641766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,2560,0.15772977140214708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,2048,0.10852711068259345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,2048,0.06788177622689141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,3072,0.19672977924346924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,2048,0.1378648943371243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,2048,0.13059022691514757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,1536,0.05738311343722873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,12288,0.7627493540445963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,2560,0.07763377825419109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,1024,0.07971288760503133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,1536,0.08474044667349921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,1024,0.04902222090297275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,768,0.045165333482954234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,1536,0.10826222101847331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,2560,0.1674106650882297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,768,0.06407733096016778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,512,0.032841778463787503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,512,0.05114755696720547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,512,0.052148444785012134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,256,0.02440888848569658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,512,0.041893333196640015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,1536,0.10517777999242146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,768,0.06482222345140246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,256,0.03958400090535482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,256,0.03788355655140347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,128,0.039464887645509504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,3072,0.18549244933658174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,64,0.019671110643280875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,1024,0.062239110469818115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,32,0.020296888218985665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,128,0.02128977742460039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,32,0.03784177700678507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,128,0.033632887734307185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,128,0.03695288962788052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,256,0.04268711143069797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,768,0.046353777249654136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,1024,0.07918222082985772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,65536,1.2404692967732747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,64,0.03844533363978068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,65536,3.114495171440972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,16384,0.6725680033365885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,51200,2.256777869330512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,51200,1.016225814819336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,16384,0.2932222154405382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,65536,3.4442310333251953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,16384,0.8881946669684516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,16384,0.7925884458753797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,12288,0.6257431242201064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,51200,3.4840791490342884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,12288,0.6433448791503906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,65536,4.540345085991754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,10240,0.4302817715538873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,10240,0.19706755214267305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,8192,0.32514845000372994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,10240,0.5448275672064887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,10240,0.48483912150065106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,8192,0.16031022866566977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,7168,0.281240012910631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,8192,0.3854684564802382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,51200,2.709007051255968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,7168,0.13835110929277208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,7168,0.37568977144029403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,12288,0.5831555790371364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,6144,0.12294933531019424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,5120,0.20661421616872153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,6144,0.2958222230275472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,8192,0.5268719991048177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,4096,0.15716532866160074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,5120,0.10549244615766738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,5120,0.24807821379767525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,6144,0.26476889186435276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,7168,0.34328267309400773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,4096,0.21693866782718232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,6144,0.3252275519900852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,4096,0.20416177643669975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,3072,0.13105066617329916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,3584,0.19167199399736193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,3584,0.18074133661058214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,3072,0.16748889287312826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,3072,0.15612355868021646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,2560,0.10444266928566827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,3584,0.08293155829111735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,2560,0.1363733344607883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,12288,0.236154662238227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,2048,0.09052355421913995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,3072,0.07579111390643649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,2048,0.11730666955312093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,2048,0.11357333925035264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,2048,0.06038133303324381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,2560,0.14303467008802626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,5120,0.2716097831726074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,1536,0.09237688779830933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,1024,0.05201511250601875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,1536,0.051854223012924194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,3584,0.14047022660573324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,1024,0.06766400072309706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,1024,0.04481688804096646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,2560,0.06929155853059557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,1536,0.08889955282211304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,4096,0.08920088741514419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,768,0.04143911268975999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,512,0.025672889418072168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,768,0.03733955489264594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,512,0.04314666655328539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,256,0.018931556079122756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,768,0.05424266722467211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,256,0.02994933393266466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,128,0.015984889533784654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,256,0.03353777858946059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,1536,0.06771377722422282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,128,0.02881422307756212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,512,0.03513333201408386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,64,0.014747555057207743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,512,0.04468977782461378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,32,0.0148062225845125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,64,0.031223111682467993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,32,0.03122755553987291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,768,0.05532088875770569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,1024,0.06782044304741754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,128,0.029094222519132826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,256,0.03587911195225186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,65536,1.0725057390001085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,65536,2.1430674658881292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,16384,0.5090151362948948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,51200,0.9074835247463651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,128,0.03262133399645487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,65536,2.815066655476888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,65536,3.3578436109754772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,16384,0.7091031074523926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,16384,0.25251110394795734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,12288,0.37819377581278485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,51200,1.9293155670166016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,16384,0.6688026852077908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,12288,0.19592621591356066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,10240,0.3510275681813558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,12288,0.5351297590467665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,12288,0.48585420184665257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,8192,0.29786398675706655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,10240,0.4157431125640869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,8192,0.14498844411638048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,8192,0.3538000053829617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,51200,2.209344016181098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,7168,0.2611315515306261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,8192,0.3280044396718343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,6144,0.2134231064054701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,7168,0.3263644377390544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,7168,0.2886417706807454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,10240,0.4375297758314345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,6144,0.10354755322138469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,51200,2.564801745944553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,6144,0.2698400020599365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,4096,0.13769066333770752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,5120,0.21050755182902017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,7168,0.12106755044725205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,4096,0.1814568837483724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,10240,0.16862222883436415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,6144,0.2656444443596734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,3584,0.13369422488742405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,4096,0.17190044456058076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,3072,0.1029848919974433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,3584,0.15994221634334987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,5120,0.22371199395921496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,3584,0.15310222572750515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,2560,0.08514755302005345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,3072,0.06655910942289564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,3072,0.1331662204530504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,4096,0.07750044266382854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,2560,0.11830311351352268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,2048,0.07142400079303317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,2560,0.11385067303975423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,3584,0.07293066713545057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,2048,0.053007112609015576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,2048,0.09718310832977295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,2048,0.09468266699049209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,3072,0.1408346626493666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,1536,0.045943111181259155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,1536,0.076273779074351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,2560,0.06099377738104927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,1024,0.05591111050711738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,768,0.027628444963031348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,1536,0.05247733328077528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,5120,0.09060000048743354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,768,0.04439111219512092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,768,0.03206933206982083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,512,0.02012977831893497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,1024,0.0384204453892178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,512,0.035728000932269625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,512,0.03751644492149353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,1024,0.05718133184644911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,512,0.026939556002616882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,256,0.02903288933965895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,1536,0.07657333215077718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,768,0.04682844546106127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,256,0.024715556038750544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,5120,0.1778471072514852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,128,0.0258568889564938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,64,0.012539555629094442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,32,0.012701333396964602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,64,0.0247013337082333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,256,0.027687110834651526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,32,0.02432800001568264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,128,0.024488889508777197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,128,0.013421333498424955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,65536,0.8559395472208658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,256,0.015232889188660515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,65536,2.3660827212863498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,65536,1.578129768371582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,16384,0.3388586574130588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,1024,0.0370506677362654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,51200,0.6754933463202583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,65536,2.12825075785319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,16384,0.2284382184346517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,51200,1.2614035076565213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,12288,0.2538462214999729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,16384,0.5001929071214464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,10240,0.22652978367275664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,12288,0.41772977511088055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,12288,0.3782782289716933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,51200,1.8492533365885417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,16384,0.5604586601257324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,8192,0.17921600076887342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,12288,0.1594408882988824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,8192,0.280257781346639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,8192,0.25655823283725315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,7168,0.14347644646962485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,10240,0.3562062316470676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,7168,0.2433626651763916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,7168,0.09575821955998738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,6144,0.14662133322821722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,128,0.02369955513212416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,6144,0.0855226649178399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,10240,0.13881422413720026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,5120,0.12659911314646402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,6144,0.19897688759697807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,10240,0.31822223133511013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,7168,0.22616267204284668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,5120,0.07468977901670668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,5120,0.1775155597262912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,5120,0.1642284393310547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,3584,0.08144444227218628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,6144,0.21073333422342935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,4096,0.13495377699534097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,51200,1.6687706841362848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,3584,0.060235553317599826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,3072,0.07782577806048922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,3584,0.11949333879682754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,4096,0.0878417756822374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,3072,0.055995557043287486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,3072,0.1104115578863356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,3072,0.10427377621332805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,4096,0.06553422080145942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,8192,0.10736978054046631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,2048,0.04978933268123203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,2560,0.09399466382132636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,2560,0.08996622429953681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,1536,0.03572355707486471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,2048,0.07765244775348239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,2048,0.07487910985946655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,1536,0.03752533263630337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,2560,0.06060622135798136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,1024,0.025429333249727886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,4096,0.14479022555881077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,2560,0.052419556511773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,1024,0.027318222655190363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,768,0.019496000475353666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,2048,0.045089777972963124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,768,0.03492177857293023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,512,0.014064888159434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,1536,0.06075288852055868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,1536,0.05971644322077433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,512,0.027415111660957336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,512,0.022515555222829182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,256,0.011776000261306763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,1024,0.043888890080981784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,1024,0.04445688923199972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,3584,0.12682044506072998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,256,0.020992888344658747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,256,0.022314666046036616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,768,0.024119999673631456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,768,0.03641422258483039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,128,0.01960355540116628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,32,0.010037333601050908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,128,0.020621332857343886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,64,0.018404444058736164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,512,0.029221332735485498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,128,0.010020444790522257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,256,0.022414222359657288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,64,0.009861333502663506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,128,0.020823111136754353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,65536,1.1123662524753146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,32,0.01791555517249637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,65536,0.8368159929911295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,51200,0.9542062547471789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,16384,0.25289689169989693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,65536,1.9749554528130426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,51200,0.6564097934299046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,16384,0.200762669245402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,12288,0.21386399534013537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,51200,1.4511697557237413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,12288,0.155257781346639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,12288,0.3470035658942328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,10240,0.1643422179751926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,65536,1.8831768035888672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,51200,1.519977781507704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,16384,0.6059279971652561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,8192,0.1410871081882053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,16384,0.4740159776475694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,10240,0.29107022285461426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,7168,0.1198915508058336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,8192,0.10630844699011908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,12288,0.35738666852315265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,8192,0.24022934171888563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,8192,0.23522043228149414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,7168,0.21045511298709443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,7168,0.20695200231340197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,6144,0.18245511584811738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,10240,0.13501866658528647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,6144,0.17873244815402559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,5120,0.07251911030875312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,5120,0.15361777941385904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,4096,0.07112533516354032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,5120,0.15112711323632136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,7168,0.09472088681326972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,3584,0.06746133168538411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,4096,0.12487733364105225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,4096,0.12327555815378825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,6144,0.08237510919570923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,10240,0.29890667067633736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,3584,0.1116595533159044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,3584,0.10922666390736897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,3072,0.05439733465512594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,2560,0.046930667426851064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,4096,0.06347999970118205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,5120,0.08991822269227769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,2560,0.08152621984481812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,2048,0.03926755653487311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,3584,0.059028446674346924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,2560,0.0820666684044732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,2048,0.04333600070741442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,1536,0.028566221396128338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,6144,0.10975555578867595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,1536,0.052687999274995595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,1536,0.05435466766357422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,1024,0.01938577824168735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,2048,0.06674222151438396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,1024,0.03783911135461595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,3072,0.09603555334938897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,768,0.015604444675975375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,1024,0.025454221500290766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,1024,0.04064177804523044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,768,0.030052443345387776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,1536,0.03324266605907016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,3072,0.09605333540174697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,512,0.024119999673631456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,2048,0.06828888919618395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,256,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,512,0.026058667235904272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,256,0.01884977850649092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,768,0.023588443795839947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,768,0.03331644336382548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,128,0.008720888859695857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,256,0.020644444558355544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,128,0.020630222227838304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,128,0.01629599928855896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,64,0.010151111417346531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,128,0.01830577850341797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,32,0.008732444710201686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,64,0.01519555515713162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,32,0.01461866663561927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,3072,0.050776001479890615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,2560,0.049514667855368726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,256,0.020279111133681405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,65536,0.8174097802903918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,512,0.021645334031846788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,65536,0.6390835444132487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,16384,0.1813733312818739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,512,0.011783999701340994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,51200,0.4998933474222819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,51200,1.2540231280856662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,16384,0.4122595522138808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,51200,0.601003540886773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,51200,1.2275706397162545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,16384,0.38872978422376847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,10240,0.12207644515567356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,12288,0.12768356005350748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,12288,0.3003733423021105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,12288,0.29200622770521373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,65536,1.638053364223904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,65536,1.5303742090861003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,10240,0.11212178071339925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,10240,0.24532087643941244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,12288,0.1341600020726522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,7168,0.08902400069766575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,8192,0.08477955394321018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,8192,0.20320444636874727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,8192,0.198707554075453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,6144,0.07560355795754327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,7168,0.17467021942138672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,8192,0.09405244721306695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,6144,0.15787466367085776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,6144,0.06731200218200684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,6144,0.15123911698659262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,5120,0.12995289431677923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,7168,0.1788782278696696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,5120,0.12751821676890054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,5120,0.059693336486816406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,10240,0.257335106531779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,16384,0.16438844468858507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,4096,0.052520000272327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,4096,0.10592177841398452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,4096,0.10394489102893406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,3584,0.04922044608328077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,7168,0.07625066571765475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,4096,0.048947556151284106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,3072,0.08137244648403592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,3072,0.0455022222465939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,3072,0.08110044399897258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,5120,0.05881510840521919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,3584,0.04488089018397861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,2048,0.025460445218616064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,2560,0.06912710931566027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,2560,0.041491554843054876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,3584,0.09268355369567871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,1536,0.019681778219011094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,2048,0.05660621987448799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,2048,0.05773244301478068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,2560,0.03247377938694424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,1024,0.014113777213626437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,1536,0.04607911242379082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,1024,0.0318213336997562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,3584,0.09483200311660767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,768,0.011785777906576792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,1024,0.02146044373512268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,768,0.02525955604182349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,1536,0.044250667095184326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,768,0.02030933399995168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,2560,0.06957511107126872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,3072,0.03875111209021674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,1024,0.03391733434465196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,2048,0.03580088747872247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,512,0.0222746663623386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,256,0.015488000379668342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,512,0.009647111097971598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,256,0.01794311073091295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,512,0.01998133295112186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,128,0.01311288856797748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,256,0.007524444825119442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,64,0.007150222029950883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,128,0.016021332807011075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,32,0.007286222444640265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,1536,0.027585777971479628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,64,0.011804444094498953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,128,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,32,0.011391999820868174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,65536,0.4610871209038629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,512,0.018851555056042142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,128,0.01789955629242791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,51200,0.3484942118326823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,65536,0.6306639777289497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,16384,0.10612622234556411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,256,0.01625333395269182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,65536,1.285172462463379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,65536,1.2302853266398113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,12288,0.08088444338904487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,51200,1.0078836017184787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,16384,0.31155021985371906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,768,0.028007109959920246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,12288,0.24650133980645073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,10240,0.06850577725304498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,12288,0.23524088329739043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,51200,0.5521759986877441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,10240,0.20725066132015654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,8192,0.05574578046798706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,16384,0.3419582313961453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,51200,0.9592889149983724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,8192,0.08318222231335111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,8192,0.16822044054667154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,8192,0.16212088531917995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,7168,0.07458844449785021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,6144,0.04638755652639601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,16384,0.15956978003184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,10240,0.19734933641221789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,6144,0.12785866525438097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,7168,0.04928000105751885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,5120,0.03553422292073568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,12288,0.12544533941480848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,5120,0.11036088731553818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,5120,0.10309955808851455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,4096,0.031339555978775024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,6144,0.06594577762815687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,6144,0.12215556038750543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,4096,0.05095200075043572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,4096,0.08435466554429795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,7168,0.14102578163146973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,5120,0.058119999037848584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,3584,0.07709955506854586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,3072,0.022416889667510986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,4096,0.08774133523305257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,3072,0.06680888599819608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,7168,0.15098488330841064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,2560,0.01702311138312022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,3584,0.02646666599644555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,2560,0.03936266567971971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,2560,0.05739555756251017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,2048,0.014294221997261047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,2560,0.056792000929514565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,3584,0.07535377475950453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,1536,0.011382222175598145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,2048,0.03015377786424425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,2048,0.047361777888404004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,1536,0.0362062222427792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,1536,0.023760888311598036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,1536,0.03719733489884271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,3584,0.047305779324637524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,1024,0.027291556199391682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,1024,0.02130399975511763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,2048,0.046668443414900035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,3072,0.06623999940024482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,1024,0.009351111120647853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,768,0.022775999373859827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,768,0.020297777321603563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,1024,0.025701334079106648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,512,0.016181333197487723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,256,0.005636444522274866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,512,0.01887822151184082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,256,0.011927111281288994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,768,0.02068711154990726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,256,0.01349511080318027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,128,0.004907555464241239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,256,0.01792622274822659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,128,0.01787911189926995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,64,0.00425866660144594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,512,0.01808088852299584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,32,0.004563555535342959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,64,0.008678221868144141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,32,0.008404444489214156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,512,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,65536,0.34320444530910915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,128,0.010025777750545079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,128,0.013804444836245643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,768,0.008015111088752747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,51200,0.26209243138631183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,65536,0.6581955485873753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,3072,0.04393511017163595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,65536,1.1602747175428603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,51200,0.5028115378485786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,65536,1.0928337309095595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,51200,0.9351723988850912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,12288,0.06654310888714261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,16384,0.16116889317830405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,16384,0.2835128837161594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,12288,0.12586222754584417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,12288,0.2136168877283732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,16384,0.08782222535875107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,8192,0.046605332030190356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,16384,0.2978168858422173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,8192,0.1583555539449056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,10240,0.055758224593268506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,8192,0.08335289027955796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,8192,0.14567644066280788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,10240,0.11190133624606663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,7168,0.13906221919589573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,6144,0.03590133455064561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,7168,0.1275742186440362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,12288,0.23183290163675943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,7168,0.04122489028506809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,6144,0.10998133156034683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,6144,0.0661697785059611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,10240,0.18958400355445015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,5120,0.09874933295779759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,7168,0.07556621895896064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,5120,0.09238400061925252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,6144,0.1205635600619846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,4096,0.0748035576608446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,4096,0.050982223616706006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,3584,0.07150844732920329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,4096,0.07987733019722833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,3584,0.06630311409632365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,5120,0.03052355514632331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,3584,0.022367111510700647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,3072,0.06087911128997803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,5120,0.05794755617777506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,2560,0.011026666396194033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,3072,0.05801066425111559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,3072,0.016545777519543965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,2560,0.050980445411470204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,2560,0.049610667758517794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,2048,0.009350222018029954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,4096,0.025342222717073228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,2048,0.04188977678616842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,3072,0.043792890177832715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,1536,0.008194666769769456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,2048,0.028727110889222886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,2048,0.041449778609805636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,3584,0.04741244514783224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,1536,0.0240328891409768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,1536,0.03236444460021125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,51200,0.8508924378289117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,1024,0.02386844489309523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,768,0.005868444426192178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,768,0.0185199992524253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,2560,0.03926755653487311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,512,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,768,0.020259555843141343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,768,0.01993955506218804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,1024,0.021163556310865615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,512,0.018616000811258953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,256,0.00425866660144594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,512,0.015840000576443143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,256,0.01791200041770935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,1536,0.03242222136921353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,256,0.012126222252845764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,1024,0.022908444205919903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,512,0.014581332604090372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,64,0.003580444388919406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,256,0.010379555324713389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,64,0.007326222128338284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,32,0.007924444145626491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,128,0.017888888716697693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,65536,0.3244497776031494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,51200,0.25124621391296387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,32,0.0038897775941424896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,16384,0.08332710795932345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,128,0.008309333688682979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,12288,0.0656026668018765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,65536,1.11444091796875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,10240,0.0542773339483473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,51200,0.8708560201856824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,12288,0.2168275515238444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,128,0.003867555409669876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,7168,0.039247999588648476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,10240,0.18254933092329237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,6144,0.03430666526158651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,7168,0.13342666625976562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,5120,0.0296106669637892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,6144,0.11240443918440078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,4096,0.024695111645592585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,1024,0.006919999917348226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,5120,0.0967893335554335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,4096,0.07591822412278917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,3072,0.014127999544143677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,3584,0.06785866949293348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,2560,0.010023111270533668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,16384,0.29593155119154185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,3072,0.05712888638178507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,2048,0.009902222288979424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,1536,0.007985778152942657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,2560,0.049047112464904785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,8192,0.04718933171696133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,1024,0.006597333484225803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,3584,0.021896888812383015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,1536,0.03035466538535224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,512,0.005133333305517833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,8192,0.15022488435109457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,256,0.00423555572827657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,512,0.013743999931547375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,128,0.0038746665749284956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,256,0.00960622231165568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,64,0.0031840000301599503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,2048,0.0396106673611535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,32,0.003546666767862108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,64,0.006595555692911148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,1024,0.022111111217074927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,768,0.017900443739361234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,32,0.007265778051482306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,65536,0.3593591054280599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,16384,0.08071821928024292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,51200,0.25928711891174316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,12288,0.06211911307440864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,128,0.007292444507280986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,10240,0.05165511038568285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,16384,0.2870559957292345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,8192,0.04338755541377597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,51200,0.8690809143914117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,12288,0.21669777234395346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,7168,0.03745955559942458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,6144,0.03305600086847941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,8192,0.15023288461897108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,7168,0.1299368937810262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,6144,0.11235111289554173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,4096,0.023355555203225877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,5120,0.09437866343392266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,3584,0.020947555700937908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,4096,0.07573333051469591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,3072,0.012773333324326409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,3584,0.06952622201707628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,5120,0.028387556473414104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,2560,0.00925066653225157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,768,0.0057546666098965546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,2048,0.007286222444640265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,3072,0.06052800019582113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,1536,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,2560,0.050796445873048574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,65536,1.115338643391927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,1024,0.005492444253630108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,768,0.004576888763242298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,10240,0.18797777758704293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,1536,0.03211199906137254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,1024,0.023034666975339253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,256,0.0033031110134389666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,512,0.014120888378885059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,128,0.0032035555276605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,256,0.009696000152164036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,64,0.003181333343187968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,128,0.007496889266702864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,2048,0.04176533222198486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,64,0.006727111008432176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,768,0.018473777506086562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,32,0.003198222153716617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,512,0.004223111189073986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,32,0.007290666302045186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,16384,11.599187215169271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,10240,0.17994488610161674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,128,0.012804444465372296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,51200,13.573732164171007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,16384,3.8288133409288196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,16384,18.307118733723957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,12288,8.053961859809027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,12288,2.9155991872151694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,12288,11.771697150336372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,16384,13.334524366590712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,12288,9.342785305447048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,10240,6.779369778103298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,10240,2.3685003916422525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,10240,8.272578769259983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,51200,33.58754136827257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,8192,1.937469270494249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,8192,5.784834543863933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,8192,5.710052490234375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,10240,7.479430304633246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,8192,5.501250796847874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,7168,1.5904275046454535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,7168,4.5191324022081165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,7168,4.4413299560546875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,6144,1.2711795171101887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,7168,4.536153581407335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,6144,4.1425395541720915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,6144,3.932575225830078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,5120,1.097090615166558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,6144,3.889124552408854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,5120,3.568639967176649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,5120,3.3645138210720487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,4096,2.7574693891737194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,4096,0.9011422263251411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,4096,2.683462142944336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,5120,3.2288693322075734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,4096,2.6469173431396484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,3584,2.1855706108940973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,3584,0.7973813480801053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,3584,2.241236368815104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,3072,0.7138524585300021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,3072,2.094095018174913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,3072,1.9079945882161458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,3584,2.300991906060113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,2560,1.7049119737413194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,3072,2.006784862942166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,2560,1.722330623202854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,2048,1.369400872124566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,2560,1.6730978224012587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,2048,0.5011582374572754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,2048,1.3289670944213867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,2048,1.3694328731960719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,1536,0.40279822879367405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,1536,1.0535857942369249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,1024,0.7394373151991102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,1536,1.0040533277723525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,1536,1.0582560433281791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,1024,0.4044622315300836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,768,0.5668026606241862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,1024,0.7866355578104655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,1024,0.7276195420159234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,512,0.4067777792612712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,768,0.5914462407430013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,768,0.5915155410766602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,512,0.5140737957424587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,256,0.2614142100016276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,512,0.48825777901543516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,2560,0.6066657702128092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,256,0.20992799599965414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,256,0.37026045057508683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,512,0.2404702239566379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,128,0.43221243222554523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,128,0.2102524439493815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,128,0.276400883992513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,64,0.1968239943186442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,32,0.19436977969275582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,128,0.21654310491349962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,64,0.4479404555426703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,32,0.4644764264424642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,768,0.35881866349114316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,256,0.45857954025268555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,51200,63.44102647569445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,65536,13.824451870388456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,51200,25.197213066948787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,51200,10.711295233832466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,65536,34.12601047092014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,16384,8.687085469563803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,16384,2.9624595642089844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,16384,14.295528835720487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,16384,10.539176093207464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,12288,6.808888753255208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,12288,2.194695154825846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,51200,37.337059868706596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,12288,9.03494601779514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,51200,49.782680935329864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,51200,45.71308051215278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,65536,46.85303751627604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,12288,7.242513020833333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,10240,1.7833287980821397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,10240,6.0117687649197045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,8192,1.3835502200656469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,10240,6.334145016140408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,10240,5.86351818508572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,8192,4.498789469401042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,8192,4.8165537516276045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,7168,3.498480902777778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,8192,4.2581562466091585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,7168,1.2152471542358398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,7168,3.7069485982259116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,6144,1.0761022567749023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,7168,3.618255191379123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,6144,3.0138195885552297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,6144,3.003029293484158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,5120,0.8649262322319878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,5120,2.4518150753445096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,5120,2.6253183152940536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,6144,3.0249404907226562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,4096,1.978700425889757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,4096,0.6816142400105795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,5120,2.5314915974934897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,4096,2.0229714711507163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,3584,0.6533333460489908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,4096,2.0068507724338107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,3584,1.8960783216688368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,3584,1.7890116373697917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,65536,64.9510498046875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,3072,1.5982489056057398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,3072,0.5378364457024468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,3584,1.7678551144070094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,3072,1.566973368326823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,2560,0.5133075714111328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,2560,1.3125360276963975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,2048,1.0698151058620877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,2560,1.3111048804389107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,2048,0.3980337778727214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,2048,1.066855960422092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,1536,0.7434533437093099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,2560,1.2434026930067275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,2048,1.0471537907918294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,1024,0.4825919999016656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,3072,1.5269163979424372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,1024,0.26251999537150067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,1536,0.7838160196940104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,1536,0.8119377560085721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,1024,0.7104017999437121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,1024,0.5519893434312608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,1536,0.3495679961310492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,768,0.46495734320746523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,256,0.2269279956817627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,512,0.3949271043141683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,768,0.4655440118577745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,512,0.370210674073961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,256,0.16596355703141955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,128,0.16630044248369005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,256,0.35908089743720156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,256,0.2885306676228841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,768,0.21868356068929037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,512,0.31884445084465873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,64,0.14804000324673122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,128,0.16774933867984346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,128,0.3381768862406413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,32,0.15225422382354736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,512,0.18647644254896376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,64,0.35145065519544816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,32,0.3638008965386285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,128,0.21892266803317598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,768,0.3936862150828044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,65536,5.029590182834201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,51200,4.027151319715712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,51200,8.424126519097223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,16384,2.948557323879666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,65536,10.209946526421442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,16384,1.1279866960313585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,16384,4.671198950873481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,16384,3.1583902570936413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,12288,2.909416834513346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,12288,2.395955615573459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,51200,11.6608395046658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,12288,2.390517340766059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,12288,0.8157591289944119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,10240,1.7310221989949544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,10240,0.7051324314541287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,10240,1.91338136461046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,10240,1.8576212988959417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,51200,16.10149468315972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,8192,1.403337796529134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,8192,0.4702613088819716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,8192,1.3783458073933919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,7168,0.40421244833204484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,8192,1.4500613742404516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,65536,15.432863023546007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,7168,1.2132061852349174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,6144,1.0171448389689128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,6144,0.3485244380103217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,7168,1.1577972835964627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,5120,0.8232577641805013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,6144,1.008771578470866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,7168,1.2377377616034615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,5120,0.8562560081481934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,4096,0.6216310924953884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,4096,0.24431999524434408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,5120,0.7211582395765516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,4096,0.5885706477695042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,5120,0.28824978404574925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,6144,0.9464826583862305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,3584,0.5657511287265354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,3584,0.2094106674194336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,3584,0.5618000030517578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,3072,0.5008826785617405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,3584,0.5164933204650879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,3072,0.4888799985249837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,2560,0.3817768891652425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,3072,0.45407464769151473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,2560,0.16904711723327637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,2048,0.3182275560167101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,2560,0.45931731330023867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,2560,0.3836355474260118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,2048,0.15348622534010145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,1536,0.26580532391866046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,3072,0.1850586732228597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,2048,0.3277884324391683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,1536,0.11792622672186957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,1536,0.26848800977071124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,4096,0.6504862043592665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,1536,0.2536071141560872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,768,0.14186222023434109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,1024,0.20059555106692842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,1024,0.18743111027611625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,768,0.0777182247903612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,2048,0.3404853343963623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,768,0.15450133217705622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,65536,22.632537841796875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,512,0.06821244292789035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,512,0.13298222753736708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,256,0.07771733072068956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,512,0.12637689378526476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,128,0.05659821960661146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,256,0.06277333365546332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,256,0.12062666151258682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,256,0.09913600153393215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,64,0.04722222354676989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,768,0.1568791071573893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,32,0.048623998959859215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,64,0.11656622091929118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,1024,0.1795591115951538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,32,0.12090578344133164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,1024,0.08556533522076076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,512,0.1065431104765998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,128,0.06342044141557482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,128,0.07715466949674818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,128,0.11266399754418267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,65536,3.861503177218967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,51200,2.9450791676839194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,51200,6.3771544562445746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,16384,2.4224480523003473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,16384,0.8877466519673666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,65536,8.796641031901041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,16384,3.5458941989474826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,16384,2.514115651448568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,51200,8.791224161783854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,12288,1.6817359924316406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,12288,0.5988453229268392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,12288,2.1912640465630426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,12288,1.6681359608968098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,10240,0.4498551156785753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,10240,1.469122674730089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,10240,1.5522480010986328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,8192,1.1062604056464302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,10240,1.3926595052083333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,8192,1.072496838039822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,51200,12.338235643174913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,8192,1.0324533250596788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,7168,0.9250151316324869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,7168,0.29673955175611705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,7168,0.9183795717027453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,6144,0.7680622206793891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,65536,10.80042690700955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,6144,0.3062986797756619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,6144,0.6635004679361979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,5120,0.5394853485955132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,7168,0.8615226745605469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,5120,0.22159022755093047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,5120,0.5499368773566352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,4096,0.5120959811740452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,65536,15.909854465060763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,4096,0.49188799328274196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,4096,0.19953155517578125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,4096,0.4464071061876085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,3584,0.4482009145948622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,3584,0.16087288326687282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,3584,0.4295830991533067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,3072,0.38406756189134383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,3584,0.40656177202860516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,3072,0.37420977486504453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,3072,0.14698044459025064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,2560,0.35482221179538304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,3072,0.3437022103203668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,2560,0.12846399678124323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,2560,0.3244311014811198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,2048,0.29161156548394096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,2560,0.29142488373650444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,2048,0.1097866694132487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,1536,0.226053343878852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,2048,0.24583644337124297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,1024,0.13804533746507433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,1536,0.19667466481526694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,1024,0.15755021572113037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,1024,0.145306666692098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,6144,0.764610661400689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,2048,0.2615404393937853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,1536,0.08769689003626506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,768,0.12171199586656357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,8192,0.33591734038458926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,768,0.11982133653428818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,768,0.06126399834950765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,512,0.055890666113959424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,512,0.10233155886332194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,256,0.05442488855785794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,5120,0.6326390902201334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,768,0.1067519982655843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,128,0.04415200153986613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,256,0.09209333525763617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,1536,0.20646399921841094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,512,0.07844355371263292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,128,0.08624444405237834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,64,0.03772977656788296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,128,0.0604648921224806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,32,0.039086222648620605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,512,0.0976488921377394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,1024,0.06871377759509616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,32,0.09311288595199585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,256,0.05145155390103658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,256,0.07664089070426093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,128,0.051057779126697116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,64,0.08902221918106079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,65536,3.3220719231499563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,51200,2.6499866909450955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,51200,5.156342400444879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,16384,1.8570914798312719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,65536,6.7028961181640625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,16384,0.6499582396613227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,16384,2.9503351847330728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,12288,1.478832032945421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,16384,2.0185813903808594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,12288,0.4393724335564508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,51200,7.603786892361111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,12288,1.7890169355604384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,12288,1.4218017790052626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,10240,1.1384151246812608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,10240,0.35426754421657985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,65536,9.188094245062933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,8192,0.8932950761583117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,10240,1.1428790622287326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,8192,0.3857742150624593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,8192,0.9182533688015408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,8192,0.8485075632731119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,51200,10.070618523491754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,7168,0.919523557027181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,7168,0.25769156879848903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,10240,1.210892465379503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,6144,0.6426142056783041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,6144,0.6314853562249078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,5120,0.5123190879821777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,6144,0.6074693467881944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,5120,0.21531289153628877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,5120,0.6244035296969944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,4096,0.41483023431566024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,5120,0.4669155544704861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,6144,0.22575110859341094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,4096,0.1850586732228597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,3584,0.33141689830356175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,4096,0.37502845128377277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,3584,0.13810843891567653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,3584,0.3668702178531223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,3072,0.2961653338538276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,3584,0.3518639935387506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,65536,14.137775844997831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,3072,0.12779377566443548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,7168,0.7802329063415527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,3072,0.29777688450283474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,2560,0.24400711059570312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,4096,0.4387199878692627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,2560,0.11102044582366943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,2560,0.2684640089670817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,2560,0.24664889441596138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,2048,0.22114488813612196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,2048,0.09438133239746094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,2048,0.20840711063808867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,3072,0.31545599301656085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,1536,0.17488622665405273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,1024,0.11935467190212673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,1536,0.16608977317810059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,2048,0.1980631086561415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,1024,0.060093336635165744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,768,0.09095644288592869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,1024,0.12483733230166966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,1536,0.15667200088500977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,512,0.06941244337293836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,768,0.10251022047466701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,1536,0.07521955834494697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,256,0.04421866602367825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,512,0.08338844113879734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,1024,0.13128711117638484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,256,0.0780586666531033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,128,0.037271999650531344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,768,0.10280711121029323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,7168,0.6961181958516439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,128,0.07311289178000556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,64,0.034625778595606484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,512,0.087281776799096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,32,0.036475555764304265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,64,0.07509066661198933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,32,0.07880355252159967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,768,0.05385955505900913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,512,0.04892444279458788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,256,0.04490310947100321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,256,0.06562844249937269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,128,0.052419556511773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,128,0.04503644506136576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,65536,2.577156490749783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,65536,4.898011525472005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,16384,1.3837884267171223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,51200,1.8905306922064886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,16384,0.49460268020629883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,16384,1.5825938118828669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,16384,2.4075334337022567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,51200,4.942031012641059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,65536,7.557515462239583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,12288,1.128303951687283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,12288,1.4591502083672419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,12288,1.071123547024197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,10240,1.027432017856174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,10240,0.29534755812750924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,12288,0.34706666734483504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,51200,8.157735188802084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,10240,0.9860497580634223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,8192,0.22740801175435385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,51200,3.9942152235243054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,8192,0.7639670901828342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,8192,0.7125768661499023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,8192,0.6930595503913032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,7168,0.23272177908155653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,7168,0.6350524690416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,65536,10.731739468044706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,6144,0.5232231352064345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,6144,0.2191502253214518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,7168,0.5555431048075358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,10240,0.8690977626376681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,6144,0.5382364590962728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,6144,0.4583582348293728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,4096,0.3300115532345242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,5120,0.6504835552639431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,5120,0.3876933256785075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,4096,0.33893778589036727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,7168,0.6149662335713705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,4096,0.312662230597602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,5120,0.3523786597781711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,3584,0.30368977122836643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,3072,0.23327467176649305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,5120,0.1507600016064114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,3072,0.1016613311237759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,2560,0.20473955737219918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,3072,0.24287199974060059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,2560,0.09113511112001206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,2560,0.22060178385840523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,2560,0.2048222223917643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,3584,0.2805173397064209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,2048,0.18219378259446886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,2048,0.07748622364468045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,1536,0.11224977175394694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,2048,0.1709208885828654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,4096,0.12625511487325033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,1024,0.08556622266769409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,1536,0.14359554979536268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,1536,0.13608800040351018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,3584,0.11441866556803386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,1024,0.10753866699006821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,768,0.06627733177608915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,1024,0.1029519968562656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,3072,0.2589493327670627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,768,0.08458933565351699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,768,0.04627911249796549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,3584,0.3089875645107693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,2048,0.16900266541375053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,512,0.07172889179653592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,256,0.03656889001528422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,512,0.06880711184607612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,256,0.06410044431686401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,256,0.054436445236206055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,1536,0.06211733155780368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,128,0.03052799900372823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,256,0.039114667309655085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,128,0.03920533259709676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,128,0.059676441881391734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,64,0.025615110993385315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,128,0.04386933313475715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,1024,0.050583998362223305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,64,0.06128978066974216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,32,0.05957333246866862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,768,0.08491377698050605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,512,0.04270222120814853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,512,0.05071466830041674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,32,0.026372444298532274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,65536,2.2674507564968533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,51200,1.6168044408162434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,51200,3.860163370768229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,65536,4.938042534722222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,16384,0.4079226652781169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,16384,1.4486817253960504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,16384,1.6191689173380535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,16384,1.906786600748698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,12288,1.2121671040852864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,65536,6.331350114610459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,51200,4.89388190375434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,10240,0.862977769639757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,12288,0.9665892918904623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,10240,0.8238968849182129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,10240,0.27521777153015137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,10240,0.7698302268981934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,8192,0.7085262404547797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,8192,0.20280088318718803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,8192,0.5998044543796116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,51200,6.939749399820964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,7168,0.5808115535312229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,65536,9.27028062608507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,12288,0.30371377203199595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,7168,0.5457297960917155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,6144,0.42158132129245335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,7168,0.49981599383884007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,5120,0.34268445438808864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,6144,0.4095715681711833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,8192,0.5933804512023926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,5120,0.3904577891031901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,12288,1.0661812888251412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,5120,0.34174222416347927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,6144,0.16050488419002956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,4096,0.11467822392781575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,4096,0.29899467362297905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,6144,0.4541671011182997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,4096,0.28179200490315753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,3584,0.10339377986060248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,3072,0.22573955853780112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,5120,0.13681244850158691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,3584,0.24578489197625053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,4096,0.31650132603115505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,3072,0.09181866380903457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,3072,0.23013244734870064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,3072,0.2180124388800727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,2560,0.1961102220747206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,2048,0.16374488671620688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,3584,0.27126312255859375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,2560,0.18213510513305664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,1536,0.11668888727823894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,3584,0.2648453447553846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,7168,0.18812978267669678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,2048,0.16246488359239367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,2048,0.1523511144849989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,1536,0.056903998057047524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,1536,0.12884444660610622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,1536,0.12204978201124404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,1024,0.09489422374301487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,768,0.07016533613204956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,2560,0.07983466651704577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,1024,0.09243910842471653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,512,0.054957333538267344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,768,0.04305244485537211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,768,0.0760542220539517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,2048,0.06783377461963229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,768,0.07671378056208293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,256,0.03966133462058173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,512,0.039043555657068886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,1024,0.08175022072262235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,2560,0.18530310524834526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,512,0.06367378102408515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,256,0.035791999763912626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,256,0.054457777076297335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,256,0.04924355612860786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,1024,0.04682133264011807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,128,0.05155466662512886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,128,0.036631110641691424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,128,0.03973689013057285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,32,0.026836444934209187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,64,0.051220443513658315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,32,0.05185244480768839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,128,0.02979822291268243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,64,0.025154666768179998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,512,0.06281422244177924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,65536,1.7384541829427083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,51200,3.0591680738661022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,16384,1.1557413736979167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,65536,4.0458221435546875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,16384,0.34600623448689777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,51200,1.3465396033393011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,16384,1.4438382254706488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,16384,1.106680022345649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,12288,0.8532124625311958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,12288,0.3061404493119982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,65536,5.643753899468316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,10240,0.6577973365783691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,51200,4.014886220296224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,51200,5.309805128309462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,12288,0.8269795311821831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,10240,0.28284621238708496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,10240,0.6725084516737195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,8192,0.6254106627570258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,10240,0.6630817519293891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,65536,7.445915222167969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,7168,0.4603084458245172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,8192,0.5339600245157877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,12288,0.8324142032199435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,7168,0.16033955415089926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,7168,0.4890186521742079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,6144,0.4352160029941135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,6144,0.1417048904630873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,5120,0.2771928840213352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,6144,0.4048062165578206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,6144,0.36400000254313153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,8192,0.2208106650246514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,4096,0.2330568896399604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,5120,0.43923200501335996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,5120,0.2986115614573161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,4096,0.26119287808736164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,4096,0.2403493457370334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,7168,0.43248534202575684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,3584,0.23138401243421766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,3584,0.09080266952514648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,3072,0.18813067012363008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,8192,0.543217764960395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,3072,0.08110577530331083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,3072,0.20111378033955893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,3072,0.18865244918399385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,3584,0.23039465480380586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,2560,0.07133599784639147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,2048,0.13810755146874323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,2560,0.1589626603656345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,3584,0.21615200572543672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,2048,0.1426533328162299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,5120,0.1236613326602512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,1536,0.10727733373641968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,2560,0.1672675609588623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,1536,0.05103200011783176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,1024,0.07245688968234591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,1536,0.11299555831485325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,1536,0.10670222176445855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,2560,0.1713555521435208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,1024,0.08342666758431329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,1024,0.08192266358269586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,2048,0.06100177764892578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,768,0.038817776574028864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,512,0.04157688882615831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,768,0.06719555457433064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,4096,0.10007199976179336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,512,0.054697778489854604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,256,0.028439998626708984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,512,0.05515911181767782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,1024,0.04338044590420193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,256,0.032856888241238065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,128,0.0244159996509552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,256,0.043465776575936206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,768,0.06736089123619927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,128,0.03303466571701898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,64,0.020670221911536325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,128,0.03581600056754218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,512,0.035340444909201726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,64,0.044621331824196704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,32,0.04508622156249153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,768,0.05797955724928114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,256,0.04783111148410373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,128,0.04508089025815328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,32,0.021336000826623704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,65536,1.4300426907009547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,2048,0.13426132996877035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,16384,0.8150666554768881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,51200,2.92338138156467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,51200,1.0794604619344075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,65536,3.944517347547743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,16384,1.0030275980631511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,16384,0.3320719930860731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,16384,0.9476497438218859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,51200,3.5633121066623263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,12288,0.6225946744283041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,51200,4.282854292127821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,12288,0.2759973208109538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,65536,5.7078950670030375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,12288,0.6956684324476453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,10240,0.5757964452107748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,65536,4.191440158420138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,8192,0.4476986461215549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,10240,0.5655537711249458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,12288,0.6877999835544162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,8192,0.18135644329918754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,10240,0.49906667073567706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,8192,0.4333413441975911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,7168,0.13886666297912598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,7168,0.390338659286499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,7168,0.3744693332248264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,6144,0.1264773342344496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,5120,0.24145865440368652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,6144,0.3044622209337023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,7168,0.36960622999403214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,5120,0.10537066724565293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,5120,0.28208356433444554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,4096,0.21583110756344268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,10240,0.19131110774146187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,5120,0.2647288905249702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,6144,0.2929813332027859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,4096,0.08819378084606594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,3584,0.17370133929782441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,4096,0.22612089580959746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,3072,0.14881243970659044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,3584,0.20230489306979707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,8192,0.4416400061713324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,3072,0.17310844527350533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,3072,0.1625022225909763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,2560,0.14748177263471815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,2560,0.13918488555484348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,2048,0.10311733351813422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,3584,0.18575999471876356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,3072,0.07099288702011108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,2048,0.12247021993001302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,2048,0.053816000620524086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,1536,0.07957333326339722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,2048,0.11814755863613552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,1536,0.046057777272330395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,6144,0.33731288380093044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,1024,0.060961776309543185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,1536,0.09363644652896458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,1024,0.03815022110939026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,2560,0.14546933439042833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,768,0.04593333270814684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,1024,0.07259200016657512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,768,0.05944177839491102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,512,0.032263110081354775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,512,0.047003553973303906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,512,0.048597332504060536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,768,0.05776888794369168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,256,0.024677332904603746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,256,0.04076000054677328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,256,0.03793244560559591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,768,0.03535733289188809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,256,0.029556443293889362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,2560,0.0633760028415256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,128,0.038633777035607234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,128,0.03137244449721442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,64,0.0196079992585712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,64,0.03806666533152262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,32,0.020534222324689228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,1536,0.09703733523686726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,32,0.03848355677392747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,128,0.029085334804322984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,4096,0.2104168865415785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,512,0.03193688723776076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,1024,0.07196799914042155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,3584,0.07963555388980441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,65536,1.0830230712890625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,65536,3.0700026618109812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,51200,2.051547580295139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,16384,0.7341733508639865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,65536,3.053529739379883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,16384,0.7765919897291395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,16384,0.2530622217390272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,12288,0.45435820685492617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,16384,0.7015760209825305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,128,0.021011556188265484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,51200,2.3712880876329208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,12288,0.18780088424682617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,12288,0.5495742162068685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,10240,0.4366790983412001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,65536,4.036111195882161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,10240,0.16587644153171116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,8192,0.31526488727993435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,10240,0.47438934114244247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,51200,3.3220782809787326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,8192,0.13640088505215114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,8192,0.3652888933817546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,8192,0.32958221435546875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,7168,0.32504354582892525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,7168,0.1190462244881524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,7168,0.29136888186136883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,12288,0.4899804327223036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,6144,0.22238577736748588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,6144,0.10212088955773248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,10240,0.41132622294955784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,6144,0.277200010087755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,51200,0.9721849229600695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,7168,0.30232445398966473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,6144,0.2500320010715061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,5120,0.09011999766031902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,5120,0.2350346777174208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,5120,0.21017777919769287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,4096,0.18678043948279488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,4096,0.17479111088646782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,3584,0.06793510913848877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,5120,0.18755822711520723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,3584,0.15296267138587102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,4096,0.1620319949256049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,3072,0.1451173358493381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,2560,0.11479200257195367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,4096,0.07544444666968451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,3584,0.15634666548834905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,2048,0.08708177672492133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,2560,0.05441688828998142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,2560,0.12362044387393528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,2560,0.11678932772742377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,3584,0.1670773294236925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,1536,0.06903200017081366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,3072,0.11938133504655625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,2048,0.10223378075493707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,1024,0.04472622275352478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,3072,0.06048533651563856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,1536,0.07945777972539265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,3072,0.13738755385080972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,1024,0.06013688776228163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,1024,0.03336266676584879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,1024,0.06063999732335409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,768,0.029775109555986192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,768,0.04840088884035746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,512,0.025872000389628943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,768,0.05036800106366476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,2048,0.09768622451358372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,512,0.026604443788528442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,512,0.04077511032422384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,768,0.0374257763226827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,256,0.033527112669414945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,128,0.01585244470172458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,256,0.0236835562520557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,2048,0.04656177759170532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,128,0.0314702226055993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,64,0.014556444353527494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,128,0.027263111538357202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,1536,0.04015555646684434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,64,0.03091111116939121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,32,0.03053155541419983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,256,0.03211911188231574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,128,0.024312888582547505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,1536,0.08114755815929837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,65536,0.9809982511732313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,512,0.0392506652408176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,65536,3.028477350870768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,51200,2.299516465928819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,256,0.019261333677503798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,65536,2.783622317843967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,65536,3.8853556315104165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,51200,0.7890782356262207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,16384,0.24290310011969674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,16384,0.6594453387790257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,16384,0.6235057512919108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,51200,2.1647680070665145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,32,0.015021334091822306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,12288,0.17913244830237496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,12288,0.4959377712673611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,10240,0.37291823493109805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,12288,0.45636622111002606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,10240,0.14789333608415392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,8192,0.29598667886522084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,10240,0.37749245431688094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,8192,0.13506133026546902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,12288,0.5907875696818033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,7168,0.2655244403415256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,8192,0.30284354421827525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,7168,0.11566310458713108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,6144,0.231711122724745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,16384,0.7028649118211535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,7168,0.27030221621195477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,10240,0.4090551005469428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,51200,3.3460985819498696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,6144,0.09267733494440715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,5120,0.20217155085669625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,6144,0.2616737683614095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,5120,0.08009511232376099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,5120,0.20946044392055937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,5120,0.19533688492245146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,4096,0.1682302157084147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,3584,0.13234933217366537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,4096,0.16082310676574707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,3584,0.14926666683620876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,3584,0.1419653362698025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,8192,0.3298444483015272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,3584,0.06241333484649658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,4096,0.06858577993181017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,3072,0.13029332955678305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,7168,0.30209867159525555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,2560,0.09418488873375787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,3072,0.12682578298780653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,2560,0.11144355932871501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,6144,0.23772621154785156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,3072,0.11384444766574436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,2048,0.09304800298478867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,2048,0.09066133366690741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,3072,0.056735999054378934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,2560,0.04977155394024319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,4096,0.1481955581241184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,2560,0.1082835528585646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,1536,0.073289778497484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,1024,0.0543137788772583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,1024,0.056180444028642446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,2048,0.04337244563632541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,2048,0.08497599760691325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,1536,0.03559199968973795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,768,0.027569777435726587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,512,0.024863110648261175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,768,0.04657155606481764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,1536,0.05832978089650472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,512,0.0354746679464976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,256,0.01963200006220076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,1536,0.07354933023452759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,1024,0.047332445780436196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,256,0.030016889174779255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,128,0.015835555063353646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,256,0.02193333374129401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,256,0.02940000096956889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,128,0.021702221698231165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,64,0.017733333839310538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,128,0.024719999896155462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,512,0.023375110493765935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,32,0.01885688967174954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,768,0.0443155566851298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,64,0.02749155627356635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,32,0.027448000179396734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,768,0.032842665910720825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,512,0.0381040010187361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,65536,0.8759057786729602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,128,0.027961777316199407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,65536,2.4878578186035156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,51200,1.8286151885986328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,16384,0.590520011054145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,51200,0.678692446814643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,65536,3.349964353773329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,1024,0.029612445169025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,16384,0.2472160127427843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,16384,0.6415831247965494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,51200,2.5862134297688804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,12288,0.37862398889329696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,51200,1.8341289096408422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,12288,0.1545626719792684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,10240,0.3244844542609321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,12288,0.43186932139926487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,65536,2.522827572292752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,10240,0.13290933767954508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,12288,0.4126888910929362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,8192,0.2791484461890327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,10240,0.37911732991536456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,8192,0.10644888877868652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,10240,0.34439555803934735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,8192,0.29145420922173393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,7168,0.09529955519570245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,8192,0.2679155667622884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,7168,0.2797955671946208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,7168,0.23874134487575957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,16384,0.5589537620544434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,6144,0.08434222141901652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,6144,0.22105688518948027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,6144,0.20418043931325278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,5120,0.18557333946228027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,5120,0.17387821939256456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,7168,0.23300533824496797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,4096,0.15032354990641275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,6144,0.20104532771640352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,4096,0.14408800337049696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,5120,0.15743288728925917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,3584,0.13355999522738987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,3072,0.09374044338862102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,3584,0.12626311514112684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,4096,0.13936000400119358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,2560,0.08917422427071466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,3072,0.05079911152521769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,3072,0.11668622493743896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,3072,0.11203111542595758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,3584,0.11535378297170003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,2048,0.06686933173073663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,2560,0.09909155633714463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,2560,0.09625955422719319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,1536,0.05275466707017687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,2048,0.0383200016286638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,2048,0.08249688810772367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,2048,0.08096355862087674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,5120,0.07290310992134942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,1536,0.06516800324122111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,1536,0.031785776217778526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,1536,0.0651493337419298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,1024,0.025961778230137292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,768,0.028033776415718928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,3584,0.05760711431503296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,768,0.03895555602179633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,768,0.023402666052182514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,512,0.02016088863213857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,768,0.042146666182412036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,4096,0.06251733170615302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,1024,0.05002133382691277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,256,0.01588355501492818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,512,0.020786666207843356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,256,0.026558222042189703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,256,0.026175111532211304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,1024,0.03266133202446832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,128,0.01310222182008955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,256,0.019298666053348117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,64,0.01218311074707243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,128,0.019256888164414298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,32,0.012492444780137805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,128,0.02235466738541921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,512,0.03127200073666043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,64,0.02439111140039232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,512,0.03407911128467984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,32,0.024246222443050806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,128,0.025123554799291823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,65536,1.6834711498684352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,51200,1.5175120035807292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,2560,0.04589244392183092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,65536,2.0519120958116317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,65536,2.6906844245062937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,51200,0.6926266882154676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,16384,0.44064799944559735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,16384,0.21163645055558947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,51200,2.028119193183051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,16384,0.5243208673265245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,12288,0.3539831108517117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,16384,0.46636178758409286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,1024,0.04804089003139072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,10240,0.3016177813212077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,12288,0.40282487869262695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,12288,0.35374755329555935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,10240,0.14312711026933458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,8192,0.23164712058173287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,10240,0.32442132631937665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,10240,0.2889537811279297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,8192,0.1045866674847073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,7168,0.21661599477132162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,8192,0.25511378712124294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,8192,0.24141600396898058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,6144,0.17341689268747965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,65536,0.7669119834899902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,7168,0.2262373235490587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,7168,0.21009155114491782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,5120,0.15008800559573704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,6144,0.07472978035608928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,6144,0.19666577710045707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,6144,0.17896533012390137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,5120,0.06623378064897326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,5120,0.16334133678012425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,5120,0.15091110600365534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,4096,0.05539377861552768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,51200,1.574453353881836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,7168,0.08640444278717041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,12288,0.1379271083407932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,3072,0.08394844664467706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,3584,0.05031377739376492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,3584,0.1164097785949707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,3584,0.11236622598436143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,3072,0.04533066683345371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,3072,0.10185244348314072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,3072,0.09757333331637913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,4096,0.13167022334204778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,2560,0.039742221434911094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,2048,0.06200889084074232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,2048,0.07191377878189087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,2048,0.07095644209120008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,2560,0.06807021962271796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,3584,0.10467377636167739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,4096,0.12713155481550428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,1536,0.02713866697417365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,1024,0.030160887373818293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,4096,0.11378933323754205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,1024,0.04282311267322964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,1536,0.05718400080998739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,1024,0.02233422299226125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,768,0.02293066680431366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,768,0.034338666333092585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,768,0.0366231103738149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,512,0.01739911072784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,1536,0.04180799921353658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,512,0.02786311176088121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,2560,0.08626577589246963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,512,0.029948443174362183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,512,0.0188711103465822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,256,0.023244443866941664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,256,0.0229111115137736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,2048,0.033486220571729876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,256,0.017749332719379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,1024,0.04405600163671705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,1536,0.05749866697523329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,128,0.02010577751530541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,64,0.01037688884470198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,128,0.017584888471497428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,32,0.01071466671095954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,64,0.021107556091414556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,32,0.020989333589871723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,768,0.020413332515292697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,128,0.012136889000733694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,256,0.013391999734772576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,65536,1.5624106725056965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,65536,0.6426409085591634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,51200,1.1323573854234483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,65536,2.1014817555745444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,51200,0.5098604626125759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,16384,0.3761395613352458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,2560,0.08441688617070515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,65536,1.7704657448662653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,16384,0.1571777794096205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,16384,0.4398800002204047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,51200,1.5534061855740016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,51200,1.3021858003404405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,12288,0.12214133474561904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,12288,0.3485111130608453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,12288,0.29365332921346027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,10240,0.10622400045394897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,10240,0.27400533358256024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,128,0.02163911031352149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,10240,0.24376445346408418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,8192,0.08162489202287462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,8192,0.21861067083146837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,8192,0.1972364452150133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,10240,0.21195022265116373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,7168,0.07350666655434503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,6144,0.13042133384280735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,7168,0.17418932914733887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,8192,0.17700710561540392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,6144,0.1677093373404609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,5120,0.11821066008673774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,6144,0.1509759955936008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,16384,0.3915528986189101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,5120,0.1410577827029758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,5120,0.12726755936940512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,7168,0.191264894273546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,4096,0.049109333091311984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,6144,0.06620711088180542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,3584,0.07357422510782878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,4096,0.10562311278449164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,12288,0.2573857837253147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,3584,0.04370933439996508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,3072,0.06995644172032674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,3584,0.0989982220861647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,5120,0.05732533004548815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,3584,0.0957440005408393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,2560,0.057413332992129855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,3072,0.038464887274636164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,3072,0.08656622303856744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,3072,0.08375733428531223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,2560,0.07388533486260308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,4096,0.1132240030500624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,2048,0.02824444572130839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,7168,0.1675706704457601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,1536,0.03686133358213637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,4096,0.09511555565728082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,1536,0.04878488845295376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,1024,0.024725332856178284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,1536,0.023751111494170293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,2560,0.03404711021317376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,2560,0.07226044601864286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,1024,0.02004888819323646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,768,0.019646222392718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,1024,0.0380968881977929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,2048,0.06150755617353651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,512,0.01459200017982059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,768,0.03176888823509216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,1536,0.04929600159327189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,512,0.023804444405767653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,2048,0.04527822136878967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,512,0.01720977822939555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,1024,0.03584622343381246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,256,0.02017600006527371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,256,0.016376889414257474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,128,0.010117332968446944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,768,0.0291884442170461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,128,0.01611555616060893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,64,0.009905777871608734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,128,0.01846311158604092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,32,0.010400888820489248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,256,0.011759999725553723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,512,0.025872000389628943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,64,0.018049778209792245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,32,0.017893332574102614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,2048,0.06133244434992472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,256,0.020963556236690942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,768,0.018592889110247295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,128,0.018360000517633226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,65536,1.090667512681749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,51200,0.7442542182074653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,16384,0.27154933081732857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,51200,0.4261653423309326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,65536,1.6141875584920247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,65536,1.455727153354221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,51200,1.0905546612209744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,16384,0.13173777527279323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,12288,0.2115386724472046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,65536,0.534038225809733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,16384,0.353749328189426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,12288,0.10514222251044379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,12288,0.27614222632514107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,12288,0.25983020994398326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,8192,0.1394533316294352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,10240,0.2224737803141276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,10240,0.21850221686893037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,8192,0.0717902249760098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,8192,0.17900178167555067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,8192,0.17712533473968506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,7168,0.12431289090050592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,10240,0.17154933346642387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,6144,0.10265688763724433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,7168,0.06375377708011203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,7168,0.15856444835662842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,5120,0.09033244185977513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,6144,0.13655822806888157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,51200,1.2458284166124132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,5120,0.050138665570153125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,4096,0.06485511196984185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,4096,0.09364444679684109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,10240,0.08956355518764919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,6144,0.13513244522942439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,4096,0.04314844475852119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,3584,0.05979022052552965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,5120,0.11425422297583686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,3584,0.0826044413778517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,3584,0.03825777769088745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,3584,0.08436977863311768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,7168,0.15634932782914904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,3072,0.07195822397867839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,3072,0.0329751107427809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,4096,0.09494844410154556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,2560,0.06155200137032402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,3072,0.052015999952952065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,2560,0.028075555960337322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,2048,0.036320000886917114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,6144,0.05652888615926107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,2048,0.05139999919467502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,2048,0.02450399928622776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,1536,0.02678577767478095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,16384,0.37675200568305117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,1536,0.04104177819357978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,1024,0.019519999623298645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,2560,0.06439733505249023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,1024,0.029858665333853826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,1024,0.03291555576854282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,2048,0.05432888865470886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,768,0.01564711166752709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,5120,0.11601421568128799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,1536,0.02130311065249973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,768,0.024756444825066462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,768,0.02815022071202596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,512,0.020017777880032856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,2560,0.04498933421240913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,256,0.010401777923107147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,512,0.01588088936275906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,768,0.01722133325205909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,1536,0.043487111727396645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,256,0.016807110773192514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,256,0.01498311095767551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,128,0.015192889504962497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,512,0.022709333234363135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,1024,0.0185199992524253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,64,0.009729777773221334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,128,0.015145777000321282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,32,0.008662222160233391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,64,0.014829332629839579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,32,0.014509333504570855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,128,0.009556444154845344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,256,0.017893332574102614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,65536,0.7567120128207736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,3072,0.07459111346138848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,51200,0.54657957288954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,65536,1.1477555168999565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,51200,0.33890310923258465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,16384,0.17388089497884116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,51200,0.9423457251654731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,51200,0.8496373494466146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,128,0.016302223006884258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,16384,0.10787110858493382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,12288,0.13661689228481716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,16384,0.2703173425462511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,10240,0.11726489331987168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,12288,0.08601333035363092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,512,0.01219911128282547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,12288,0.20970222685072157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,65536,1.0766631232367623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,8192,0.09407555394702488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,10240,0.07451377974616157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,10240,0.17649511496225992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,10240,0.1733093394173516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,8192,0.14174132876926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,16384,0.28205956353081596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,8192,0.14062578148312038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,7168,0.05276977684762743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,7168,0.1254862282011244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,6144,0.07086310784022014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,7168,0.12495733631981744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,12288,0.20549333095550537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,5120,0.05728799766964383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,6144,0.046959112087885536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,6144,0.10849955346849229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,6144,0.10798133081860012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,4096,0.04864977796872457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,5120,0.09282844596438938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,5120,0.09263288974761963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,4096,0.07389955388175116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,4096,0.07602310842937894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,3584,0.0660053359137641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,7168,0.08356177806854248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,3072,0.03525866733656989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,3584,0.02972888946533203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,3584,0.0680053300327725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,4096,0.03482666611671448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,3072,0.05734133058124118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,2560,0.029440889755884807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,3072,0.059736000166998975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,2560,0.022591999835438196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,2560,0.04890222350756327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,2048,0.024442666106753882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,2560,0.05247199866506788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,8192,0.05805955330530802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,1536,0.01959466603067186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,2048,0.020136000381575692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,2048,0.040708445840411715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,2048,0.04331733451949226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,1024,0.014523555835088095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,1536,0.032645334800084434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,3584,0.038703110482957624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,3072,0.025648888614442613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,768,0.011980444192886353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,1024,0.015983111328548856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,1024,0.0243848893377516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,1024,0.0270480000310474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,768,0.015063110325071545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,768,0.019880000087949965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,768,0.022991110881169636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,512,0.016249777542220224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,256,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,512,0.018587556150224473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,1536,0.03525688913133409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,256,0.013482666677898832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,128,0.006663110935025745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,512,0.009363555245929295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,65536,0.41597867012023926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,128,0.011810666984981961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,128,0.013311111264758639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,512,0.014041778114106921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,128,0.014498665928840637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,32,0.007294221884674496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,64,0.011729777687125735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,32,0.011389333340856763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,256,0.013167111410035027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,256,0.01575288838810391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,65536,0.5478222105238173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,5120,0.04129688938458761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,51200,0.43250756793551975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,1536,0.01806400054030948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,16384,0.14892978138393825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,51200,0.3288622167375353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,65536,0.9245298173692492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,51200,0.7172880172729492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,16384,0.10630044672224258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,16384,0.23457598686218262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,65536,0.4049235449896918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,12288,0.08364355564117432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,65536,0.9654453065660266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,12288,0.17808800273471406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,51200,0.7798595428466797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,10240,0.07339733176761203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,16384,0.24697955449422201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,10240,0.15041422843933105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,8192,0.07157511181301542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,64,0.007285333342022366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,8192,0.05715022484461466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,8192,0.12670844131045872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,12288,0.18440177705552843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,8192,0.12217511071099176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,10240,0.08931555350621541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,6144,0.05513866742451986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,7168,0.05180089010132683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,7168,0.11059733231862386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,5120,0.0440959996647305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,7168,0.10796355538898045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,6144,0.045653333266576133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,10240,0.15388088756137425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,6144,0.09610311190287273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,12288,0.10606933302349514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,4096,0.03515911102294922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,5120,0.04033599959479438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,5120,0.08075200186835395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,5120,0.07986222373114692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,4096,0.033089776833852134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,4096,0.06559289164013334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,4096,0.06627644432915582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,3072,0.028088887532552082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,3584,0.05810133616129557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,3584,0.05871022409862942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,3072,0.05056800113783943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,3072,0.02552088929547204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,2560,0.024632889363500807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,6144,0.09394311242633396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,3584,0.030730668041441176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,2560,0.04333422250217862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,2048,0.019259555472267997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,2560,0.04494933287302653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,3584,0.028986665937635634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,2048,0.020033778415785897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,2048,0.03791822327507867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,1536,0.02866222129927741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,3072,0.051967998345692955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,1024,0.012200000385443369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,1536,0.017948443690935772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,2560,0.022163555026054382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,768,0.010127999716334874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,1024,0.01590844492117564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,1024,0.021536888347731695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,1024,0.023391111029518977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,512,0.007704888780911763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,768,0.017669333351982962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,768,0.0200311111079322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,512,0.014465777410401238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,512,0.0138568886452251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,256,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,512,0.016708445217874315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,1536,0.03066844410366482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,128,0.005640000104904175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,7168,0.06601688596937391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,256,0.011896889242861005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,256,0.01348977784315745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,64,0.00795999997191959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,128,0.010386666489972008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,1536,0.016786666380034555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,128,0.013450666434235044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,64,0.010023111270533668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,32,0.009697777529557547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,65536,0.401077323489719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,256,0.013170666992664337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,65536,0.37181777424282497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,128,0.013231111069520315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,768,0.014861333701345654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,65536,0.8297582202487521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,16384,0.10699466864267985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,51200,0.2520737648010254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,51200,0.6461520195007324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,51200,0.6015413072374132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,12288,0.07634933127297296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,16384,0.08448177576065063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,2048,0.03622311022546556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,12288,0.06737777921888563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,12288,0.15429866313934326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,10240,0.06160710917578804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,12288,0.15071911282009548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,10240,0.05980000231001112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,16384,0.21065955691867402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,8192,0.048248890373441905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,10240,0.1268577840593126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,16384,0.19788800345526802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,51200,0.3021928999159071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,7168,0.04299555553330315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,8192,0.045444445477591626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,8192,0.10728177759382461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,6144,0.04014844364590115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,7168,0.04124088750945197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,7168,0.09459822045432197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,7168,0.09190666675567627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,5120,0.03186310993300544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,6144,0.0816755559709337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,6144,0.07955911424424914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,32,0.005247999810510212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,5120,0.032292445500691734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,4096,0.024319999747806128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,8192,0.10370844602584839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,65536,0.7644364568922254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,3584,0.023062222533755835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,4096,0.024554666545655992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,4096,0.05639377567503187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,10240,0.12969066037072077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,3584,0.02164622147878011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,3072,0.019204444355434842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,3584,0.05032799972428215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,5120,0.0676293306880527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,3072,0.042697777350743614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,2560,0.016961778203646343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,3072,0.044495999813079834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,2560,0.017616000440385606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,2560,0.03659910957018534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,6144,0.03696088989575704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,2560,0.03863022062513564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,3584,0.04898488852712843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,2048,0.03052977720896403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,1536,0.011749332977665795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,5120,0.06809155808554755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,1536,0.024254222710927326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,4096,0.055418666866090566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,2048,0.01460800071557363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,1024,0.01828888886504703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,2048,0.03171199891302321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,3072,0.01925155520439148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,768,0.00797333319981893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,1024,0.02032888929049174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,1536,0.026313778426912095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,768,0.015403555499182807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,512,0.006603555546866522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,768,0.01756266587310367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,512,0.012144000165992312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,512,0.012587555580668978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,512,0.014488889111412896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,1024,0.013479111095269522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,256,0.01240799989965227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,128,0.004603555632962121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,768,0.012823999755912356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,1536,0.01477777792347802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,128,0.011481777661376528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,128,0.00850755555762185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,256,0.005290666802061929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,32,0.004562666846646203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,64,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,256,0.010048888623714447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,32,0.008010666403505538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,256,0.011755555868148804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,1024,0.00905955582857132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,51200,0.19474577903747559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,65536,0.31572622723049587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,128,0.011797332929240333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,65536,0.6637671258714464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,16384,0.06268977456622653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,51200,0.2491208977169461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,65536,0.2450817690955268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,51200,0.4728399912516276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,16384,0.17227999369303384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,16384,0.08334133360120986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,16384,0.15835466649797228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,64,0.0042408886882993905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,10240,0.037616888682047524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,12288,0.13087199793921575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,12288,0.12078666687011719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,65536,0.6033244662814671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,8192,0.03226577904489305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,10240,0.11054222451315986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,51200,0.5204150941636827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,12288,0.04734311170048184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,7168,0.027492443720499676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,2048,0.01608533329433865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,7168,0.08026578028996785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,10240,0.05926489167743259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,7168,0.04064711266093784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,10240,0.10236444738176133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,8192,0.08954399824142456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,8192,0.08368800083796184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,5120,0.016787555482652452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,6144,0.06480889187918769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,12288,0.06640977991951837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,5120,0.03153511219554477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,5120,0.05514488948716057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,6144,0.023000000251664057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,7168,0.07435111204783122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,4096,0.02387199964788225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,4096,0.046217779318491616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,3584,0.013152889079517789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,4096,0.046262221203909985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,6144,0.036031110419167414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,3072,0.011745777395036487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,3584,0.021242666575643752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,3584,0.040968888335757785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,5120,0.057052446736229785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,2560,0.01038577738735411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,3072,0.03611911005443997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,3072,0.01921066641807556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,3072,0.03619822197490268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,2048,0.008932444784376357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,2560,0.017813333206706576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,8192,0.04462577899297079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,6144,0.0693768858909607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,1536,0.007663110891977946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,2048,0.025429333249727886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,2048,0.02641955514748891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,3584,0.04116355710559421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,1024,0.0063973334100511335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,1536,0.020439111524158053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,1536,0.021740444832377966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,1024,0.015669332610236276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,768,0.005584888988071018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,2560,0.03102488981352912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,2560,0.03091199861632453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,2048,0.016143111719025504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,768,0.01311911145846049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,768,0.012813333008024426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,768,0.01482399966981676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,1536,0.014859555496109856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,512,0.01072533345884747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,256,0.0042151109211974675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,512,0.012496888637542725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,1024,0.013471110827393003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,1024,0.01684888866212633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,128,0.0038924444880750445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,256,0.008378666308191087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,256,0.010728889041476779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,512,0.004937777916590373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,64,0.003666666646798452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,128,0.010384889112578498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,32,0.003589333345492681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,4096,0.014797333214018079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,512,0.012067555553383298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,32,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,65536,0.18179110685984293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,128,0.011448000040319232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,65536,0.5892551210191515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,65536,0.32028889656066895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,65536,0.5883804427252876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,16384,0.045736889044443764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,51200,0.25450311766730416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,51200,0.4617439905802409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,51200,0.46387375725640195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,16384,0.15422399838765463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,16384,0.08437422249052261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,128,0.007167999943097432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,16384,0.15294222036997476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,10240,0.030517333083682593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,12288,0.11738844712575276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,64,0.006673777682913675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,12288,0.11735555860731338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,8192,0.02538577715555827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,10240,0.09875555833180745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,256,0.011396444506115384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,8192,0.04629244407018026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,12288,0.03735288977622986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,7168,0.022602667411168415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,8192,0.0811777777141995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,12288,0.06748000118467543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,7168,0.04179822074042427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,51200,0.14025510682000056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,7168,0.07197778092490302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,10240,0.05958577659395006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,5120,0.011730666789743634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,10240,0.09922399785783555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,6144,0.06276355849372016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,8192,0.07977866464191012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,4096,0.009703111317422655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,5120,0.05345333284801907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,7168,0.07002488772074382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,4096,0.04446399874157376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,3584,0.009107555780145857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,5120,0.05072888731956482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,6144,0.015508444772826301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,3584,0.03951377669970194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,3072,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,4096,0.024528889192475214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,6144,0.060652441448635526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,3072,0.01997244523631202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,2560,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,3072,0.03456977672047085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,3584,0.0363031095928616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,2560,0.01835466590192583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,2048,0.0069475554757648045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,4096,0.0411208901140425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,3584,0.021319111188252766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,6144,0.03734577695528666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,1536,0.00600622221827507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,2048,0.025233777032958135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,3072,0.0318853325313992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,5120,0.03167111012670729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,1536,0.020970667401949566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,1024,0.004974222017659081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,1536,0.015113777584499784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,1024,0.014175110393100314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,1024,0.013506666653686099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,2048,0.022851554883850947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,2048,0.016524444023768108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,1024,0.016336888074874878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,768,0.011876444849703046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,768,0.014531556102964612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,1536,0.01841600073708428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,512,0.009719111025333405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,512,0.012080888781282636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,2560,0.027116444375779893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,512,0.01180266671710544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,768,0.004569777597983678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,256,0.011407111254003314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,256,0.0075697774688402815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,256,0.010036444498433007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,768,0.013078221844302284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,128,0.011446221835083432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,64,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,32,0.0034293333689371743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,64,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,2560,0.030276444223192003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,32,0.00599911105301645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,65536,0.17879111236996123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,128,0.0032115555885765287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,51200,0.133623997370402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,16384,0.05196266704135471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,512,0.004212444441186057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,128,0.006302222195598815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,65536,0.5676977899339464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,10240,0.02944799926545885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,51200,0.4414275487263997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,16384,0.15145333607991537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,256,0.0037884443170494507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,7168,0.021374222305085924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,10240,0.09182755814658271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,6144,0.015441776977645027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,8192,0.07627911037868924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,5120,0.01033866653839747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,7168,0.06526133086946276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,4096,0.009031111167536842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,6144,0.05674311187532213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,12288,0.03561866614553664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,3584,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,3072,0.0076044441925154785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,2560,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,4096,0.03874222106403775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,12288,0.10988089111116196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,3584,0.034670223792394005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,1536,0.00563822231358952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,2560,0.025927111506462097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,1024,0.004622222234805425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,1536,0.017811555001470778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,768,0.004571555389298333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,1024,0.013747555514176687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,512,0.003863111138343811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,5120,0.04865333437919617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,256,0.003556444413132138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,768,0.011459555890825061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,2048,0.006312888943486744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,128,0.0031795555518733132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,512,0.009371555513805812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,256,0.007140444384680853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,128,0.005734222216738596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,32,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,65536,0.1754453314675225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,8192,0.025044444534513686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,3072,0.030171556605233088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,51200,0.12757866912417942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,64,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,16384,0.048523555199305214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,51200,0.44077422883775497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,64,0.0028977776981062363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,10240,0.028175999720891316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,2048,0.02193333374129401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,8192,0.023447111248970032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,10240,0.091921779844496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,7168,0.020848888489935134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,8192,0.07434578074349298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,16384,0.146579556994968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,6144,0.013763555222087435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,5120,0.009349333743254343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,7168,0.06526755624347262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,4096,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,6144,0.05689689185884264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,3584,0.007326222128338284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,65536,0.5780826674567329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,3072,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,3584,0.03684622380468581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,2560,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,12288,0.03628622161017524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,3072,0.031553778383466936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,2048,0.005406222409672207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,1536,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,2560,0.027217777238951787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,1024,0.004575110971927643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,12288,0.10963822073406643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,5120,0.047934220896826855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,1536,0.018246221873495314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,512,0.0038782221575578055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,4096,0.03910311063130697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,256,0.003354666754603386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,768,0.011747555600272285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,128,0.0031928889867332247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,512,0.009473777479595607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,64,0.0028728888266616394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,256,0.007349333001507654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,32,0.0031840000301599503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,128,0.005663110978073544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,2048,0.02258666687541538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,768,0.004233777936961916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,32,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,1024,0.01387288918097814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,32,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,64,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,16384,1.9217910766601562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,12288,4.529279073079427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,16384,5.917693244086371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,16384,6.736678229437934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,12288,1.4944435755411785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,12288,4.692789289686415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,51200,7.415195041232639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,10240,1.1531093385484483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,10240,3.7243423461914062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,8192,0.9086613125271268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,10240,3.5957794189453125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,8192,2.9298985799153647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,8192,2.8274292416042752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,7168,0.7908097902933756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,7168,2.5767555236816406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,7168,2.389507505628798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,6144,2.3767172495524087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,6144,0.7047831217447916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,6144,2.1437742445203996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,5120,0.5879866811964247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,5120,1.5343333350287542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,5120,1.7335973315768773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,4096,0.4863484700520833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,4096,1.6004204220241969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,4096,1.3663333257039387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,3584,0.4328631030188666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,128,0.009746666583749983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,3584,1.1878408855862088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,3072,0.3827839957343207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,3072,1.1016071107652452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,3072,1.016225814819336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,2560,0.3376257684495714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,2560,1.0096089045206706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,3584,1.1827679740058052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,2560,0.8581528663635254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,51200,19.03886244032118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,2048,0.7034951315985786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,1536,0.5527040163675944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,2048,0.7567040125528971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,1024,0.413008001115587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,2048,0.28329510158962673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,768,0.2805413405100505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,1536,0.22670578956604004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,768,0.15673422813415527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,512,0.21014132764604357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,768,0.3453751140170627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,256,0.1476222276687622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,512,0.13505244255065918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,512,0.27958401044209796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,128,0.12064888742234971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,1024,0.4125395615895589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,256,0.2537084420522054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,64,0.10319466723336114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,128,0.12029777632819282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,1536,0.6483075353834364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,32,0.10946222146352132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,128,0.23839823404947916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,64,0.24521867434183756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,32,0.25004532602098256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,1024,0.17619289292229545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,256,0.11989955107371013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,16384,4.288030412462023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,51200,5.661908467610677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,51200,29.859737820095486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,65536,7.774807400173611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,16384,1.3786853154500325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,16384,4.913018544514974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,12288,1.0039386749267578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,12288,3.438966327243381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,10240,0.8764702479044596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,12288,3.6506114535861545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,10240,2.9443147447374134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,51200,13.974169413248697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,8192,0.696782218085395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,8192,2.2384605407714844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,8192,2.5593768225775824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,10240,2.8444035847981772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,7168,1.9761182996961806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,7168,1.9409075842963324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,6144,1.6625501844618056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,6144,1.5977342393663194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,6144,0.5387911266750759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,65536,17.647554185655384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,5120,0.4560400115119086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,5120,1.3870080312093098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,4096,0.3956444528367784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,7168,0.6303431193033854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,5120,1.3451022042168512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,3584,0.9104346169365777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,3584,0.3423164420657688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,3072,0.30194134182400173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,4096,1.1700693766276042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,4096,1.0504675971137154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,3072,0.8047937817043729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,2560,0.27458132637871635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,2560,0.6725200017293295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,2560,0.7471555603875054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,3584,0.9496008555094401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,3072,0.841682646009657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,2048,0.22018044524722627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,51200,21.341033935546875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,1536,0.17809689044952393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,2048,0.5509929127163357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,1024,0.31066666709052193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,1536,0.43441067801581484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,1024,0.13979110452863905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,768,0.12171733379364014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,2048,0.5114977624681261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,768,0.2739066547817654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,512,0.10832533571455215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,512,0.22061778439415825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,1536,0.5456444422403971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,256,0.09530222415924072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,768,0.23163112004597983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,256,0.1992968850665622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,512,0.1821919944551256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,64,0.0817306637763977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,128,0.09637600183486938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,128,0.18768266836802164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,32,0.08218666579988268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,256,0.12709421581692165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,64,0.19329244560665557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,32,0.19672800434960258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,128,0.09443555275599162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,1024,0.3373573356204563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,51200,1.6703661812676325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,65536,2.4505182902018228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,16384,0.46608532799614805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,16384,1.3818640179104273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,51200,4.6586558024088545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,12288,1.184003512064616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,65536,5.541505601671006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,12288,1.125473764207628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,12288,0.3488008975982666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,10240,0.2931742138332791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,10240,1.0253075493706596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,16384,1.6902728610568578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,8192,0.7811964352925619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,8192,0.33904798825581867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,10240,0.9242915047539605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,8192,0.7226293351915148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,7168,0.23294133610195586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,51200,6.821463267008464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,6144,0.5719768736097548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,6144,0.18192977375454375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,65536,32.70055813259548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,5120,0.4576133092244466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,6144,0.5323466724819607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,5120,0.15629333919949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,4096,0.3741351233588324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,5120,0.5032240019904243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,4096,0.12991021739112005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,7168,0.6795111232333713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,3584,0.32333776685926646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,3584,0.11545777320861816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,3072,0.23365510834587944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,3072,0.10319377978642781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,3584,0.3527528974745009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,65536,9.173917134602865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,7168,0.7101937929789225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,2560,0.09016533030403985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,2560,0.2262995507982042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,2048,0.07882933484183417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,2048,0.19922933313581678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,1536,0.06398133436838786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,1024,0.0916737781630622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,1536,0.1474062204360962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,4096,0.382787545522054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,768,0.07440977626376681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,1024,0.11109511057535808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,2560,0.23005599445766875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,768,0.04672799838913811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,3072,0.2650390995873345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,2048,0.16299111313290066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,256,0.04275733232498169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,512,0.042509333954917065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,1536,0.1480275524987115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,512,0.0764791104528639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,256,0.03847466574774848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,64,0.028246220615175035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,128,0.038322667280832924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,32,0.02921244502067566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,128,0.06440533532036676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,1024,0.05195022291607327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,64,0.06610577636294894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,32,0.06723733080757989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,768,0.08851555320951675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,128,0.03254666593339708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,256,0.06846844487720065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,512,0.06237600247065226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,16384,1.0269003974066842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,65536,1.7851058112250433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,51200,1.2708720101250541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,16384,0.35321421093410915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,16384,1.1927191416422527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,12288,0.9072915183173286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,51200,3.192794587877062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,12288,0.8802613152398003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,10240,0.7097128762139214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,10240,0.2230879995557997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,8192,0.4908000098334418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,12288,0.26951556735568577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,8192,0.1808533271153768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,65536,4.638824039035373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,7168,0.44183821148342556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,7168,0.16109600332048205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,6144,0.455028427971734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,7168,0.47739378611246747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,6144,0.14077422353956434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,5120,0.3327377902136909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,10240,0.7220951186286079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,51200,5.393917507595486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,5120,0.33946755197313094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,4096,0.27282844649420845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,4096,0.10309688912497626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,3584,0.2461306783888075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,65536,6.714762793646918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,3584,0.09098844395743476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,3072,0.20693066385057238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,3584,0.2354959911770291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,3072,0.08247733116149902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,2560,0.16533511214786106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,3072,0.20358310805426705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,2560,0.07157688670688205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,2048,0.1421697802013821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,2560,0.1733857790629069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,2048,0.06048533651563856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,2048,0.14357154899173313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,6144,0.42069599363538956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,1024,0.07013689147101508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,1536,0.0498942236105601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,1536,0.11388266086578369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,768,0.06302666664123535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,8192,0.5791573524475098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,1024,0.08586311340332031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,512,0.043567111094792686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,768,0.03799555699030558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,4096,0.29940978686014813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,256,0.031067556805080835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,1536,0.0996026661660936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,256,0.031537777847713895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,256,0.05303288830651177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,5120,0.13380355305141872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,128,0.031601776679356895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,128,0.0498595568868849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,1024,0.04156622290611267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,32,0.023264888260099623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,64,0.05106222298410204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,32,0.05244444476233589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,512,0.035086221165127225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,128,0.026456000076399908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,64,0.022613333331214056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,768,0.06889777713351779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,512,0.05910311142603556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,65536,1.408725314670139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,16384,0.9322702619764539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,16384,0.39712177382575137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,16384,0.9984381993611654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,51200,2.9279787275526257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,12288,0.6790275573730469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,12288,0.22665511237250435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,65536,3.819646199544271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,10240,0.5910551283094618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,10240,0.19187733862135145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,8192,0.4774888886345758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,10240,0.6874355740017362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,51200,4.174020555284288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,7168,0.4383760028415256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,8192,0.6641297870212132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,12288,0.7483102480570475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,7168,0.39763111538357204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,51200,1.062103059556749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,6144,0.41893866327073837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,5120,0.2711733447180854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,6144,0.3438284397125244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,8192,0.15331110689375135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,4096,0.2155075603061252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,5120,0.28637422455681694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,4096,0.08486666944291855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,3584,0.18862577279408774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,4096,0.22449866930643717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,3072,0.17680088678995767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,3584,0.1996204455693563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,7168,0.13726754983266196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,3072,0.07232533560858832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,2560,0.13449954986572266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,3072,0.17259644137488472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,2048,0.11981333626641168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,5120,0.13363288508521184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,65536,5.6492114596896705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,2560,0.1472746663623386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,2048,0.05127911104096306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,2048,0.1223564412858751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,1024,0.06496977806091309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,1536,0.04333066609170702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,768,0.04670755399598015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,3584,0.0761680006980896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,1024,0.07417688767115276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,768,0.033688889609442815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,512,0.03361777795685662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,768,0.05945955382453071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,256,0.026392888691690233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,2560,0.061562663979000516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,512,0.050640887684292264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,1536,0.09681422180599636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,128,0.02253066665596432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,256,0.04515377680460612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,64,0.019168888529141743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,1536,0.09722933504316543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,128,0.04246933261553446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,32,0.019174221489164565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,6144,0.12351289060380723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,32,0.044496887260013156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,256,0.02790400054719713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,128,0.0280693338976966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,1024,0.03662933243645562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,64,0.04373688830269707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,512,0.03103111187616984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,65536,1.3369742499457464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,16384,0.6708408991495768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,51200,1.01219908396403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,65536,2.7518897586398654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,16384,0.2677680121527778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,51200,2.142862319946289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,12288,0.19993155532413057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,12288,0.5942168765597874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,10240,0.49045599831475156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,8192,0.382055123647054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,8192,0.37634844250149196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,51200,3.744940439860026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,7168,0.3059324423472087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,65536,4.773693508572049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,10240,0.16525422202216253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,7168,0.11576355828179254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,6144,0.28173777792188853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,6144,0.10089155700471665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,8192,0.13422932889726427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,16384,0.8308515548706055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,5120,0.09057422478993733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,4096,0.18400711483425566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,4096,0.07501777675416735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,7168,0.35460623105367023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,3584,0.1651137802335951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,5120,0.22771022054884169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,3584,0.06668355729844835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,3584,0.16392444239722356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,12288,0.6516302426656088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,2560,0.11701156033409967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,5120,0.23500800132751465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,3072,0.14197422398461237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,2048,0.09342666467030843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,4096,0.18630399968889025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,2560,0.12144088745117188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,1536,0.07179733117421468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,2048,0.04412088791529337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,2048,0.10075200266308254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,3072,0.15039288997650146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,10240,0.4659137725830078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,1536,0.08057866493860881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,3072,0.05853688716888428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,1024,0.031207998593648274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,1024,0.06075111362669203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,768,0.02803022331661648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,2560,0.051338666015201144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,768,0.048766222265031606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,256,0.021497777766651575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,512,0.025218667255507574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,6144,0.27864710489908856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,128,0.01817599932352702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,256,0.023388443721665278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,1024,0.05023555623160469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,64,0.015222221612930298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,128,0.023217777411142986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,32,0.01552888916598426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,128,0.03400444322162204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,64,0.03413688805368211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,32,0.03496799866358439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,512,0.041486223538716636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,1536,0.03678399986690945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,768,0.03715822100639343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,512,0.031173331869973078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,65536,1.1365199618869357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,16384,0.6219875547620985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,51200,0.8305608961317273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,65536,2.5406977335611978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,16384,0.24195377031962076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,51200,2.107023027208116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,16384,0.7908870908949109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,12288,0.5493377579583062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,12288,0.17659466796451143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,10240,0.4492008950975206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,10240,0.1850524478488498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,12288,0.5799359745449489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,256,0.03650488787227207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,10240,0.43355732493930393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,8192,0.1401537789238824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,7168,0.29894667201571995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,7168,0.10343022478951348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,51200,3.059334225124783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,8192,0.46486224068535703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,7168,0.2906862099965413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,6144,0.09079911311467488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,65536,4.277640024820964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,6144,0.2476257748074002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,4096,0.15777599811553955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,5120,0.20874310864342582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,3584,0.1204951074388292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,8192,0.3530986573961046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,4096,0.18404355314042833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,3072,0.1205635600619846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,3584,0.14411644140879312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,6144,0.22348443667093912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,3072,0.053575111760033496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,5120,0.20551910665300158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,5120,0.07880000273386638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,2560,0.04689688815010918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,2048,0.08305511209699842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,4096,0.06715022193060981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,1536,0.06504622432920668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,3584,0.05905422237184313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,1536,0.03345866666899787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,1536,0.07168622149361505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,1024,0.04471910993258158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,3072,0.12653511100345188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,2560,0.09918044673071967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,768,0.03716177741686503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,1024,0.028025776147842407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,512,0.025426665941874187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,768,0.02496888902452257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,2560,0.10873244206110637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,768,0.04373066624005636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,256,0.01959022217326694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,2048,0.04001511136690775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,512,0.03601955705218845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,256,0.021165332860416834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,256,0.03136266602410211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,64,0.014512888259357877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,128,0.02954933378431532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,32,0.014420444766680399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,64,0.029667556285858154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,32,0.030010667112138536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,512,0.02249777813752492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,128,0.016549333930015564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,128,0.021104888783560857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,1024,0.0539644459883372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,2048,0.09002044465806748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,65536,0.9174879921807183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,16384,0.5509769121805826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,65536,2.07313961452908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,16384,0.2717679871453179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,12288,0.4101102352142334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,51200,1.8754143185085719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,16384,0.6180791325039333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,12288,0.16194844245910645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,10240,0.3454897668626573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,12288,0.45105600357055664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,10240,0.13042578432295057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,8192,0.29374223285251194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,51200,2.775901370578342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,10240,0.5191502041286892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,7168,0.26588532659742564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,8192,0.3102160029941135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,7168,0.09432799948586358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,6144,0.22796355353461373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,65536,3.466947555541992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,5120,0.16939200295342338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,5120,0.07145600186453925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,5120,0.1821813318464491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,51200,0.7148550881279839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,4096,0.059522668520609535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,3584,0.11400266488393147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,3584,0.052593777577082314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,7168,0.2871662245856391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,3072,0.09831555684407552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,6144,0.21734400590260824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,2560,0.08983288870917426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,3072,0.04730044470893013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,3072,0.11114488707648383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,2560,0.041672888729307384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,8192,0.10208355718188816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,2560,0.09523733456929524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,2048,0.03518133362134298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,1536,0.050309333536359996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,2048,0.07987022399902344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,6144,0.08032711346944173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,1536,0.029119110769695703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,4096,0.12914844353993735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,1536,0.06333600150214301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,1024,0.024588444166713293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,1024,0.04747200012207031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,512,0.021817778547604878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,768,0.038718223571777344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,4096,0.14503910806443956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,256,0.017247110605239868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,512,0.019641778535313077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,512,0.03168355425198873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,128,0.015167999598715039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,256,0.01844888925552368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,256,0.027435556054115295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,64,0.01759111053413815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,768,0.028245333168241713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,128,0.026084444589085046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,768,0.02187999917401208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,64,0.025755556093321905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,2048,0.06941866874694824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,32,0.025747555825445387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,3584,0.12722844547695583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,128,0.018246221873495314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,32,0.018279999494552612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,65536,0.7822977701822916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,65536,1.7354675928751628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,51200,0.5868133438958062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,16384,0.6348355611165365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,1024,0.03749422232309977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,16384,0.1876595550113254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,12288,0.3639529016282823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,12288,0.1282151142756144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,16384,0.5683982107374403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,10240,0.3045413229200575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,65536,2.7338532341851125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,51200,2.2166737450493708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,10240,0.11241243945227729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,51200,1.5269102520412867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,10240,0.3105422125922309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,7168,0.22329778141445586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,8192,0.25844534238179523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,6144,0.15601778030395508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,7168,0.21554133627149794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,6144,0.07085066371493869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,12288,0.36933422088623047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,6144,0.18641244040595162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,4096,0.11650666925642227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,5120,0.06026311053170098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,8192,0.2588595549265544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,5120,0.1564275556140476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,4096,0.12495644887288411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,3584,0.044996443721983165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,3072,0.08951289123959011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,7168,0.07923377884758843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,2560,0.06924444437026978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,5120,0.152547558148702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,2560,0.0351235568523407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,2048,0.056001775794559054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,4096,0.05088266730308533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,8192,0.09261422024832831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,1536,0.043641779157850474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,2048,0.06854222218195598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,3072,0.04022488991419474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,1024,0.03175910976197984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,1536,0.054413331879509814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,1024,0.021746666895018682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,2560,0.08249155680338542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,768,0.02350577712059021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,3584,0.09674400091171265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,512,0.018556444181336295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,768,0.019496889577971566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,768,0.033654222885767617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,256,0.014116444521480136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,3584,0.10920711358388265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,512,0.027263111538357202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,128,0.012429333395428128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,256,0.016509332590632968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,256,0.023574221465322707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,128,0.016571554872724745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,128,0.022673777408070032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,1024,0.04104355639881558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,64,0.022478222846984863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,2048,0.030851556195153132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,32,0.02234755622016059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,512,0.017601778109868366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,1536,0.02534666657447815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,64,0.01074399964676963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,32,0.01074399964676963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,16384,0.36895378430684406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,51200,0.4752995702955458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,65536,1.6189147101508246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,3072,0.09614488813612197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,16384,0.16035466723971897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,12288,0.3168106608920627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,16384,0.42679111162821454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,12288,0.1276924477683173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,10240,0.251784880956014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,51200,1.6929324467976887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,12288,0.3009271091885037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,8192,0.19481955634223092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,51200,1.2296568552652996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,8192,0.07628799809349908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,7168,0.16673599349127877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,8192,0.2177333301968045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,6144,0.14332266648610434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,7168,0.1782826715045505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,6144,0.07025688886642456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,5120,0.11959644158681233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,65536,0.5893644226921929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,65536,2.149897681342231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,4096,0.0871502227253384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,5120,0.05195822318394979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,10240,0.09282222059037949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,3584,0.08247199985716078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,4096,0.04213066564665901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,10240,0.2540568775600857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,4096,0.10469600227144028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,3072,0.06848444541295369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,3584,0.037189334630966187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,2560,0.05512711074617174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,7168,0.06734933455785115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,2048,0.0465191106001536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,2560,0.06833777825037639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,6144,0.15324444240993923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,2048,0.025549333956506517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,1536,0.03579466541608175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,5120,0.13933422830369738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,1536,0.021886222892337378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,1536,0.045906666252348155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,3584,0.09314222468270196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,3072,0.08198933468924628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,1024,0.018625777628686693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,768,0.020262221495310467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,1024,0.03423733181423611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,512,0.014834667245546976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,768,0.017085333665211994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,2048,0.05720711416668362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,768,0.02830844455295139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,512,0.01572088897228241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,512,0.02328444520632426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,256,0.014815110299322339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,256,0.020006222857369315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,1024,0.02440711193614536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,3072,0.033979554971059166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,64,0.010305778019958073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,128,0.01482222146458096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,128,0.018935999936527677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,64,0.018632888793945312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,32,0.018931556079122756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,256,0.011753777662913004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,128,0.010762666662534079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,65536,0.5604773097568089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,32,0.010447999669445885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,65536,1.3155706193712022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,16384,0.3314809004465739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,16384,0.16289867295159235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,12288,0.24344444274902344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,51200,1.411243544684516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,12288,0.11702222294277614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,12288,0.3209831184811062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,2560,0.029812445243199665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,10240,0.08576266633139716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,10240,0.23794844415452746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,8192,0.06894844770431519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,16384,0.35983022054036456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,8192,0.18285600344340006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,65536,1.8917164272732203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,7168,0.0602417786916097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,10240,0.2146364450454712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,51200,0.8396453327602811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,8192,0.15237066480848524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,51200,0.4543760087754991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,5120,0.10415199730131362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,6144,0.13873244656456843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,4096,0.07700800233417086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,5120,0.04649510979652405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,5120,0.11669422520531549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,7168,0.14663110838996038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,4096,0.038266665405697294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,4096,0.09506044122907852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,3072,0.057024889522128634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,3584,0.03442488776312934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,6144,0.12836000654432508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,3584,0.08425066868464152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,3072,0.0303422212600708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,2048,0.041704889800813466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,2560,0.027179555760489568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,6144,0.05363377928733826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,1536,0.03198933270242479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,2048,0.052377776967154614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,3584,0.0666524436738756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,1024,0.024999999337726172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,1536,0.04186133212513394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,7168,0.15956089231703016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,2560,0.047135998805363975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,1024,0.031073778867721558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,3072,0.07357955641216703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,512,0.015657777587572735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,2560,0.06253688865237765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,768,0.025292444560262892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,2048,0.023523555861579046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,256,0.012072888513406118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,512,0.014815110299322339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,512,0.021037333541446265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,256,0.0140684445699056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,1536,0.020103999310069613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,256,0.018241778016090393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,64,0.008303999900817871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,128,0.013955555028385587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,128,0.01722400055991279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,768,0.020080889264742535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,64,0.017024889588356018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,32,0.016885333591037326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,768,0.01605599953068627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,128,0.00962577760219574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,65536,0.4622062047322591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,32,0.008471999731328752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,65536,1.1654266781277127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,1024,0.017242666747834947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,51200,0.959869384765625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,51200,0.36606044239468044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,16384,0.12983466519249812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,12288,0.21595022413465711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,12288,0.08744444449742635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,51200,1.173781288994683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,16384,0.3148168987698025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,12288,0.30480533175998265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,8192,0.14168532689412436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,10240,0.20179643895890978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,8192,0.05928711096445719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,7168,0.12659911314646402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,8192,0.16514666875203451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,7168,0.05384177631802029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,7168,0.141047994295756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,65536,1.5393760469224718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,10240,0.18644355403052437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,6144,0.12313955359988743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,10240,0.07684533463584052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,5120,0.040729777680502996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,5120,0.10479644272062515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,16384,0.2768186728159587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,4096,0.033598221010631986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,3584,0.05748533540301853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,3584,0.02975999977853563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,3072,0.05309244328074985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,6144,0.04787733488612705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,2560,0.04208533300293816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,3072,0.02712000078625149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,3072,0.06502044200897217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,2048,0.03462844424777561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,2560,0.055733336342705615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,6144,0.11124444007873535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,1536,0.027719999353090923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,2048,0.046979556481043495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,3584,0.07486844725079007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,1536,0.01870044403605991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,5120,0.09593066904279922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,1536,0.03754577702946133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,768,0.015671110815472074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,4096,0.06851910882525973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,1024,0.02774311105410258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,512,0.012127999630239276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,768,0.02314222190115187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,4096,0.08501333660549587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,256,0.009687110781669617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,512,0.019187556372748483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,256,0.013056000073750814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,1024,0.02053600052992503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,1024,0.016241777274343703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,128,0.015484443969196744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,2560,0.024362666739357844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,64,0.008472888833946651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,32,0.008671999805503422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,2048,0.021495110458797876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,64,0.015457777513398064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,32,0.015476443701320224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,256,0.01627466744846768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,65536,0.42605866326226133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,512,0.013833777772055732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,65536,1.2734959920247395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,128,0.009005332986513773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,16384,0.2505857679578993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,51200,1.0190515518188477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,16384,0.11745244926876491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,16384,0.27721065945095485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,12288,0.19112888971964517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,51200,0.6472400029500326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,12288,0.08031377527448866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,12288,0.21120266119639078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,10240,0.06963111294640435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,8192,0.13080444600847033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,10240,0.17585955725775826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,7168,0.11356266339619954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,768,0.01493866741657257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,8192,0.14225155777401394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,7168,0.0483822226524353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,7168,0.1244586706161499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,10240,0.1687999963760376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,6144,0.041909333732393056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,6144,0.10679377449883355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,51200,0.3400106694963243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,4096,0.06500889195336236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,5120,0.09133955505159165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,128,0.013152889079517789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,3584,0.057320886188083224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,6144,0.09548533625072902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,4096,0.07361688878801134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,3072,0.05132533444298638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,3584,0.026524444421132404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,3584,0.06435733371310763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,2560,0.043403555949529014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,3072,0.05635999970965915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,5120,0.0351217786471049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,2560,0.021935999393463135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,2560,0.04855644371774462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,4096,0.02941333254178365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,2048,0.019162666466501024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,1536,0.027448000179396734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,2048,0.0409804450141059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,1024,0.020004444652133517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,1536,0.016935999194780987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,5120,0.08344088660346137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,1536,0.03299999899334378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,3072,0.024081778195169237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,1024,0.024821332759327356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,8192,0.05583910809622871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,768,0.01391377713945177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,2048,0.03471022182040744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,512,0.012936000194814471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,256,0.009043555292818282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,768,0.01422933406300015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,1024,0.014846222268210517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,256,0.014653333359294467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,128,0.008037333687146505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,256,0.012778666284349231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,64,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,128,0.012341332932313284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,128,0.013836444252067141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,32,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,64,0.013479111095269522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,512,0.01702844434314304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,65536,0.6238231129116482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,65536,0.3386720021565755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,768,0.020443555381562974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,65536,0.8561333550347222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,51200,0.4941733148362901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,16384,0.17552267180548775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,65536,1.0207768546210396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,51200,0.27414488792419434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,16384,0.08684977557924058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,16384,0.23365955882602266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,12288,0.06756799750857882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,51200,0.802154646979438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,12288,0.18040711349911162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,10240,0.058247109254201256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,10240,0.14842132727305093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,512,0.011730666789743634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,8192,0.045816000964906484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,8192,0.12149066395229763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,6144,0.07456088728374906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,7168,0.040672888358434044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,12288,0.13725066184997559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,7168,0.10531110896004571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,5120,0.05801600217819214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,10240,0.1241439978281657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,6144,0.09193155500623916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,5120,0.030229333374235366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,5120,0.07775200075573392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,3584,0.04314666655328539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,4096,0.0628053347269694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,7168,0.07556711302863227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,3072,0.03335288829273648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,3584,0.05527999997138977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,32,0.013559111290507846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,2560,0.029987557066811457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,3072,0.021370665894614324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,3072,0.048672000567118325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,4096,0.04740000102255079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,2560,0.019668444991111755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,8192,0.1040488878885905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,1536,0.019996444384257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,2048,0.035440001222822405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,1536,0.015507555670208402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,1536,0.028446220689349707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,6144,0.035197334157096014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,1024,0.021486222743988037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,1024,0.01351288871632682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,2560,0.041871110598246254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,4096,0.025648888614442613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,512,0.00941244430012173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,768,0.018273777431911893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,3584,0.02336799932850732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,1024,0.014584889014561972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,512,0.011794666449228922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,512,0.015211555692884656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,256,0.011412444214026133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,128,0.0069511110583941145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,256,0.012763555679056378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,768,0.011973333027627734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,64,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,128,0.011431111229790581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,32,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,128,0.012072888513406118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,768,0.012771555946932899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,32,0.011782222323947482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,256,0.007305777735180325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,65536,0.28635554843478733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,51200,0.4135884443918864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,2048,0.024567999773555334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,16384,0.14839111434088811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,65536,0.8003910912407769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,2048,0.017250667015711468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,12288,0.11640355322096084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,51200,0.6358124415079752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,16384,0.19507377677493623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,10240,0.08872444762123956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,12288,0.06263377931382921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,65536,0.4929493268330892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,12288,0.147526224454244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,10240,0.05551377932230631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,10240,0.1246231132083469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,7168,0.06382222308052911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,8192,0.043502221504847206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,8192,0.10100799798965454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,7168,0.03780000077353583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,7168,0.08859911229875352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,16384,0.08505333132214016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,64,0.011731555892361535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,5120,0.04432177874777052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,8192,0.07636888821919759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,5120,0.06444533003701104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,6144,0.05577422512902153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,4096,0.0529386666085985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,3584,0.0333555539449056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,4096,0.02463200026088291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,3584,0.022224000758594934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,6144,0.07708622349633111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,3072,0.026724444495307073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,4096,0.03714488943417867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,2560,0.02327288852797614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,3072,0.019907555646366544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,3072,0.041570666763517596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,2048,0.019928889142142404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,2560,0.018428444862365723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,2560,0.03601955705218845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,2048,0.016521778371598985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,2048,0.03030222323205736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,1024,0.011838222543398539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,1536,0.013845333622561561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,3584,0.04730133215586344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,1536,0.02478755513827006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,1024,0.012725333372751871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,1024,0.018598222070270114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,512,0.00797333319981893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,768,0.015501333607567681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,51200,0.22634755240546334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,256,0.006503110958470239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,512,0.011378666592968835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,512,0.01312888910373052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,128,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,256,0.01074488874938753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,256,0.011066666907734342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,128,0.010644444160991244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,768,0.009861333502663506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,128,0.010394666757848527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,32,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,768,0.011874666644467248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,64,0.010198222266303169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,32,0.010392000277837118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,65536,0.3780959977044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,1536,0.015750222735934787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,65536,0.22124711672465006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,51200,0.3053617742326525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,16384,0.0961831079588996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,6144,0.03302933441268073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,51200,0.18134755558437773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,65536,0.6124044524298774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,12288,0.07593511210547553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,16384,0.05950044261084663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,10240,0.053469333383772105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,12288,0.04791733291414049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,8192,0.052298665046691895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,12288,0.11750843789842393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,10240,0.09928711255391438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,7168,0.04313155677583483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,8192,0.08038489023844402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,7168,0.027300443914201524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,7168,0.0707119968202379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,6144,0.037096887826919556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,5120,0.02831288841035631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,51200,0.4783448643154568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,5120,0.030136889881557886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,6144,0.061392002635531955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,10240,0.04236444499757555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,5120,0.021141333712471858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,5120,0.051389333274629384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,8192,0.03131999903255039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,4096,0.04262577825122409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,3584,0.016784888174798753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,3072,0.020081778367360432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,16384,0.18659999635484484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,6144,0.023945776952637568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,2560,0.01680266691578759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,3072,0.015445333388116626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,4096,0.025546666648652818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,2048,0.014819555812411837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,2560,0.014207111464606391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,3584,0.022749332918061152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,2560,0.029083556599087183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,1536,0.01202044470442666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,2048,0.013111111190583972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,1024,0.009363555245929295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,1536,0.01198933356338077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,3584,0.03824088970820109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,1536,0.02030044462945726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,768,0.008028444316652086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,1024,0.011058666639857821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,512,0.0069413334131240845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,1024,0.015375999940766228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,768,0.010611555642551847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,768,0.012890666723251343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,512,0.010057777994208867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,512,0.011066666907734342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,4096,0.017941332525677152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,256,0.009362666971153682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,2048,0.024713777833514746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,64,0.005535999933878581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,128,0.008683555656009251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,128,0.009364444348547194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,32,0.004808889081080754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,256,0.005909333212508096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,3072,0.03401511245303684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,32,0.00832088871134652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,128,0.00490399988161193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,65536,0.3278915617201063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,65536,0.215101334783766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,51200,0.2381217744615343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,65536,0.5222080018785265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,16384,0.0840275552537706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,51200,0.17250666353437635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,12288,0.059398220645056836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,16384,0.059697780344221324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,16384,0.13446844948662653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,12288,0.047361777888404004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,12288,0.1025164458486769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,8192,0.04175733195410835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,10240,0.08632089032067193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,64,0.004563555535342959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,7168,0.03834844297832913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,8192,0.029863112502627905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,7168,0.02684177789423201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,7168,0.061830222606658936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,256,0.009692444569534725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,6144,0.02387911081314087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,6144,0.05327111151483324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,5120,0.04500177833769056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,5120,0.020643555455737643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,4096,0.021000888612535264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,10240,0.0414720012081994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,5120,0.024824000067181055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,3584,0.017652443713612027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,4096,0.03722933265897963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,8192,0.06954044765896268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,3584,0.016653334101041157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,3072,0.01588355501492818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,3072,0.015339555011855232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,6144,0.028444445795483057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,2560,0.013450666434235044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,2560,0.014333332578341166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,2048,0.011416888899273343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,4096,0.017976888351970248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,2048,0.013063111239009433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,51200,0.41457245084974503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,3584,0.033041778537962176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,1536,0.012081777883900536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,10240,0.048671109808815845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,1024,0.007627555893527136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,3072,0.029341333442264136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,768,0.006776888751321369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,1024,0.011085333095656501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,1024,0.013908444179428948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,512,0.0058151111006736755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,768,0.011735999749766456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,1536,0.009885333478450775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,512,0.010018666585286459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,256,0.004568888909286923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,256,0.009715555442704095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,1536,0.01791822248035007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,256,0.008343111309740279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,128,0.009568889107969072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,2560,0.025592888394991558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,128,0.007688889073001013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,32,0.003923555629120933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,64,0.007654222349325816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,512,0.010032888915803697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,32,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,65536,0.22378312216864693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,128,0.0042275554604000514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,65536,0.17152888245052764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,64,0.0041617775956789655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,51200,0.1369102266099718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,16384,0.06057688925001356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,51200,0.34657955169677734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,12288,0.04195466637611389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,16384,0.04707644383112589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,16384,0.11441866556803386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,10240,0.038126223617129855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,12288,0.03815999958250258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,12288,0.08709510829713608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,51200,0.17416621579064262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,8192,0.02723466687732273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,10240,0.07332711087332831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,65536,0.44003110461764866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,768,0.010710222025712332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,8192,0.059102223979102246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,6144,0.019015999303923715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,7168,0.020728000336223178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,64,0.00832088871134652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,7168,0.051891555388768516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,6144,0.018477333916558158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,6144,0.04558488726615906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,4096,0.014183999763594734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,5120,0.016712889075279236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,5120,0.03882844580544366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,8192,0.023561777340041265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,4096,0.01477777792347802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,3584,0.01276088919904497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,4096,0.03212622139188979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,2048,0.021697777840826247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,5120,0.016759110821617972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,3584,0.028191109498341877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,3072,0.025416889124446448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,3072,0.012862222062216865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,10240,0.034035556846194796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,2048,0.008652444514963362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,2560,0.012117333710193634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,7168,0.023248889380031165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,1536,0.007630222373538547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,2048,0.018749333090252347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,3072,0.011443555355072021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,1024,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,3584,0.013791110780504016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,2560,0.01035733355416192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,768,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,1024,0.010025777750545079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,1024,0.012086222569147745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,512,0.004906666775544484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,768,0.009700444837411245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,2048,0.011711111499203576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,256,0.0042408886882993905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,512,0.009024000002278222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,1536,0.01075644459989336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,128,0.0038782221575578055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,256,0.007334222396214803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,64,0.003629333443111844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,128,0.008714666797055138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,32,0.0035911111368073356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,128,0.006739555547634761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,64,0.006657777975002925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,2560,0.02199199961291419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,65536,0.17805066373613146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,768,0.010391111175219217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,512,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,51200,0.13875644736819798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,65536,0.1666168901655409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,1536,0.015620445211728414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,65536,0.35536887910630965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,51200,0.2871608999040392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,16384,0.04732266730732388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,16384,0.09445777866575454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,10240,0.02932888931698269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,12288,0.03775022096104092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,12288,0.07282400131225586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,8192,0.021112889051437378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,32,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,10240,0.05963022179073758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,7168,0.014685332775115967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,8192,0.02368800010946062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,8192,0.049192888869179614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,51200,0.13429066869947645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,6144,0.012882666455374824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,12288,0.03496444556448195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,7168,0.042916443612840444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,6144,0.018932445181740653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,6144,0.037280887365341187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,4096,0.010315555665228102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,5120,0.031996445523367986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,10240,0.03432000014517043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,3584,0.009020444419648912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,4096,0.014845333165592618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,4096,0.02626222206486596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,16384,0.0456853343380822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,3072,0.008310221963458592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,3584,0.013975110318925647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,256,0.008708444734414419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,3584,0.023689778314696416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,5120,0.011423110961914062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,2560,0.01236977759334776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,2560,0.01849244369400872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,5120,0.0169831116994222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,1536,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,2048,0.011754666765530905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,2048,0.015846222639083862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,1024,0.004659555438492033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,1536,0.010963555839326648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,1536,0.01330044451687071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,768,0.004584889031118817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,2560,0.007303111255168915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,7168,0.021261332763565913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,512,0.004208000169859992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,3072,0.020970667401949566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,2048,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,256,0.003539555602603488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,512,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,512,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,128,0.003516444315512975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,256,0.00630044440428416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,128,0.00870311094654931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,128,0.00599644457300504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,1024,0.010616000327799054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,32,0.0032257777121331957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,768,0.009379555781682333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,768,0.009000889129108852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,65536,0.09886844290627374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,32,0.005924444645643234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,256,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,65536,0.17212355136871338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,65536,0.3249697685241699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,51200,0.0795724458164639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,1024,0.010032888915803697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,16384,0.03155466583040025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,3072,0.013106666505336761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,51200,0.13668266932169595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,12288,0.027894222074084814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,16384,0.0487973325782352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,51200,0.24778400527106392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,64,0.005648889061477449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,12288,0.03940000136693319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,12288,0.06433510780334473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,10240,0.03571555680698819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,10240,0.05309777789645725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,7168,0.012117333710193634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,8192,0.04378133349948459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,64,0.0032097777972618737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,6144,0.01072355525361167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,10240,0.01661688917213016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,6144,0.03331199950642056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,5120,0.009575111170609793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,6144,0.019710222880045574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,8192,0.024493333366182115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,4096,0.00796088907453749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,5120,0.017639999588330586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,3584,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,4096,0.01531022290388743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,7168,0.022495999932289124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,7168,0.0381368895371755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,3072,0.006940444310506185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,3584,0.014303111367755465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,2560,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,3584,0.021373333202468023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,3072,0.013495999905798169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,2048,0.005554666535721884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,16384,0.08370933267805312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,2560,0.01679733395576477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,2048,0.011744000017642975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,2048,0.014648889501889547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,5120,0.028519110547171697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,1024,0.004253333227501975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,1536,0.012442666623327466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,4096,0.023756444454193115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,768,0.003942222230964237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,1024,0.009840000006887648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,3072,0.01902755598227183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,8192,0.01349866638580958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,512,0.003583999971548716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,1536,0.00519822206762102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,512,0.00905777762333552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,512,0.007294221884674496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,1536,0.011056889262464313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,256,0.008676444490750631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,256,0.005976000179847081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,1024,0.010025777750545079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,128,0.008719111482302347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,128,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,32,0.002982222164670626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,2560,0.012429333395428128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,768,0.008432889150248634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,64,0.005304000029961268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,65536,0.09137511253356934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,256,0.0034222222036785553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,32,0.005376889059940974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,16384,0.031048887305789526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,12288,0.024658666716681585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,128,0.003181333343187968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,51200,0.24709510803222656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,16384,0.08312266402774386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,8192,0.013951111170980664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,12288,0.06408622529771593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,7168,0.012071111136012606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,10240,0.05308977762858073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,6144,0.010762666662534079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,8192,0.04314488834804959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,5120,0.010384889112578498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,7168,0.03801777627733018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,4096,0.008153777983453538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,6144,0.03291555576854282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,3584,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,5120,0.028361777464548748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,10240,0.015656888484954834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,3072,0.0069955554273393415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,4096,0.023382221659024555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,2560,0.006602666858169768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,64,0.0028524444335036804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,2048,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,3584,0.021207110749350652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,1536,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,3072,0.018969777557584975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,2560,0.016703110602166917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,768,0.003924444317817688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,1536,0.012166221936543783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,512,0.003543111185232798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,1024,0.009677333136399588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,256,0.003206222214632564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,768,0.008423999779754216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,65536,0.3154364426930745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,128,0.0028799999919202593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,512,0.007329777710967594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,64,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,256,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,32,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,128,0.005299555758635203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,1024,0.004559999952713649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,64,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,65536,0.09172266721725464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,32,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,51200,0.07490577962663439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,16384,0.02846933404604594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,12288,0.02094399929046631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,65536,0.31651467747158474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,51200,0.24704265594482422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,16384,0.08269066943062676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,8192,0.010750222537252637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,10240,0.05283111002710131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,7168,0.01019377758105596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,2048,0.014469333820872836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,51200,0.07322221994400024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,6144,0.009737778041097853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,5120,0.008676444490750631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,7168,0.037759111987219915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,4096,0.008385777473449707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,6144,0.03311555584271749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,10240,0.012707555459605323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,12288,0.06398133436838786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,3072,0.008638222184446123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,2560,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,768,0.009731555150614845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,2048,0.005221333354711533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,3072,0.019606222709019978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,1536,0.004565333326657613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,2560,0.017349332571029663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,1024,0.004020444634887907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,2048,0.01459111107720269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,768,0.0036319999231232535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,1536,0.012429333395428128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,512,0.003524444583389494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,1024,0.009714666340086196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,256,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,768,0.008456889126035903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,128,0.002833777831660377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,512,0.007303111255168915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,64,0.0025315555847353404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,8192,0.04301955633693271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,32,0.0027751111321979095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,128,0.005285333428117964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,64,0.005249777601824866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,32,0.0053102220926019884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,3584,0.007751999629868402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,4096,0.023697778582572937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,3584,0.021686222818162706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,256,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,5120,0.028338667419221666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,16384,1.8248365190294054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,16384,5.288685268825955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,16384,6.220755683051215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,12288,4.197241041395399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,16384,5.738262176513672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,51200,7.135526021321614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,12288,1.3428257836235895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,12288,4.291302151150174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,10240,1.0912222332424588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,12288,4.049334208170573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,10240,3.7260089450412326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,10240,3.7752168443467884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,8192,0.8367102411058215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,10240,3.294089847140842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,8192,2.975375917222765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,8192,2.758912828233507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,7168,0.7657377984788684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,51200,17.17724778917101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,8192,2.5942560831705728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,7168,2.4226195017496743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,6144,0.6650977664523655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,7168,2.3261627621120877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,7168,2.633395512898763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,6144,2.0403707292344833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,6144,2.085423151652018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,5120,0.5555733574761285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,5120,1.6937413745456273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,5120,1.699503156873915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,4096,1.421303113301595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,5120,1.7038692898220484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,6144,2.029217825995551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,4096,0.4649590916103787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,4096,1.3728613323635523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,3584,0.41232977973090273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,3584,1.2222169240315754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,4096,1.2792658276028104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,3584,1.165007061428494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,3072,0.9450604120890299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,3584,1.160087161593967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,3072,0.3674151102701823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,2560,0.7877377933926053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,3072,1.0303999582926433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,2560,0.3202151192559136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,3072,1.022165298461914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,2560,0.8507413334316678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,2048,0.6216044425964355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,2560,0.8293253580729166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,2048,0.2710026635064019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,2048,0.6972008811102973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,1536,0.4702489111158583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,2048,0.6795262230767144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,1024,0.3356462319691976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,1536,0.5111635526021322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,1024,0.444900459713406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,1024,0.17093866401248506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,1024,0.3897262149386936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,768,0.24936355484856498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,1536,0.21865599685245088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,1536,0.5741724438137478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,768,0.15347554948594835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,512,0.2051831086476644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,51200,27.2118157280816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,768,0.3082311153411865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,256,0.1525217824512058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,512,0.131315549214681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,512,0.27746666802300346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,128,0.11463022232055664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,256,0.11944711208343506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,256,0.198543111483256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,256,0.25785599814520943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,64,0.09574311309390598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,128,0.2362888918982612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,768,0.32438932524787056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,64,0.24248978826734754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,32,0.2474595440758599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,512,0.2498373455471463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,128,0.11650044388241237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,128,0.15092178185780844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,32,0.09993955824110244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,65536,7.203764173719619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,51200,13.295740763346354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,51200,5.4432890150282125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,65536,17.124681260850696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,16384,4.204969618055555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,16384,1.3458631303575304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,16384,5.385713789198134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,51200,20.657943725585938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,16384,4.441006978352864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,12288,3.2443262736002603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,12288,1.0844106674194336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,12288,3.724961598714193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,12288,3.239958233303494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,10240,2.799272961086697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,10240,0.8558177947998047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,65536,27.642030504014755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,10240,2.842780431111654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,65536,21.458343505859375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,10240,2.6376541985405817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,8192,2.1809510125054254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,8192,2.215700361463759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,7168,2.053733401828342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,8192,2.1234238942464194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,7168,0.609790219200982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,7168,1.9387707180447047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,6144,1.5610017776489258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,6144,1.5883768929375544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,6144,0.5421902338663737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,8192,0.6761013136969672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,5120,1.2975439495510526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,5120,0.44164178106519913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,5120,1.322721799214681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,7168,1.8129022386338975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,4096,1.03002134958903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,5120,1.302001741197374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,4096,0.3689688841501872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,4096,1.0609244240654838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,3584,0.9926053153143989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,4096,1.009834713406033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,3584,0.3240319887797038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,3584,0.9514008627997504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,3072,0.2957271205054389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,3584,0.8814497523837619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,3072,0.8597848680284289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,3072,0.7961555586920844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,3072,0.7706266509162055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,6144,1.5496400197347004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,2560,0.6896808942159017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,2560,0.26386043760511607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,2560,0.6772124502393934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,2048,0.543265766567654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,2048,0.2215342256757948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,1536,0.38884443706936306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,2048,0.5450000233120388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,2048,0.5812622176276313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,1536,0.17336532804701063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,1536,0.517265796661377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,1024,0.29468266169230145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,1536,0.41724443435668945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,1024,0.13543910450405544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,1024,0.33153245184156627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,1024,0.3024764325883654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,2560,0.6234320004781088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,768,0.257695992787679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,512,0.15838221708933511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,768,0.24259021547105578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,512,0.21862310833401152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,768,0.22159022755093047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,256,0.12390489048428005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,768,0.12159821722242568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,256,0.1973066727320353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,256,0.1571520037121243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,256,0.09238221910264756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,512,0.10523111290401882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,128,0.18621244695451525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,128,0.12004889382256402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,64,0.07622577746709187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,32,0.0791226691669888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,64,0.19093778398301867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,128,0.09193599886364406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,32,0.1947502162721422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,512,0.19843022028605142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,128,0.09275555610656738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,65536,2.323765436808268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,51200,4.273767259385851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,65536,6.105708652072483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,51200,1.6852009031507704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,16384,1.524521827697754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,51200,7.190514458550347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,65536,6.227402581108941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,16384,0.4348302152421739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,65536,9.288441128200954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,51200,19.297154744466145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,12288,1.154553731282552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,16384,1.723306655883789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,16384,1.4432267083062067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,12288,1.1355946858723958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,10240,0.9151333702935113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,12288,1.0053475697835286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,10240,0.9256631003485786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,51200,4.9112904866536455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,10240,0.8024871084425184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,8192,0.2293191221025255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,12288,0.3558639950222439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,8192,0.7257635328504773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,8192,0.6065422164069282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,7168,0.6712924639383951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,7168,0.20496533976660836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,7168,0.6195768780178493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,7168,0.5448942184448242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,6144,0.4997982449001736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,8192,0.8330666753980848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,5120,0.42842933866712785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,6144,0.4609164661831326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,10240,0.28212179077996147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,5120,0.46512799792819554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,4096,0.341830227110121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,4096,0.12552177906036377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,6144,0.17842843797471789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,4096,0.3467457824283176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,6144,0.5927697817484537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,3584,0.2869653436872694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,5120,0.1562053362528483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,3584,0.11574133237202962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,5120,0.4023537900712755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,3072,0.10326577557457818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,3072,0.2621884346008301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,3072,0.25001690122816295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,2560,0.21696889400482178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,2560,0.0920613341861301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,2560,0.2232524553934733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,2560,0.21216089195675322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,3584,0.27766044934590656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,2048,0.07337688737445407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,2048,0.18458844555748832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,1536,0.1423733366860284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,2048,0.17506578233506945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,1536,0.06167111131880018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,1536,0.14612444241841635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,1536,0.14111377133263484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,2048,0.176511115498013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,1024,0.10984000232484604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,1024,0.10526844527986313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,4096,0.31614579094780815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,768,0.08789333369996812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,1024,0.09891200065612793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,768,0.08726488881640965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,512,0.056523554854922824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,1024,0.05014311273892721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,512,0.07561155822541979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,768,0.06949688990910848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,512,0.07119022475348578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,512,0.04171466827392578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,256,0.03741066654523214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,256,0.06768888897365995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,128,0.032369779215918645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,256,0.05618933174345228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,3584,0.3044186698065864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,768,0.04598755637804667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,128,0.03718311256832547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,128,0.06382755438486735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,128,0.04526666800181071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,64,0.06527911292182074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,256,0.03831555445988973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,32,0.06654755274454753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,3072,0.2522515455881755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,64,0.028034667174021404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,32,0.0292577776643965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,65536,1.6705155902438695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,65536,4.2922647264268665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,51200,3.3833830091688366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,51200,1.255864037407769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,16384,1.1858089235093858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,16384,1.2274444368150499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,16384,0.37391021516587997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,16384,1.0439875920613606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,51200,5.734515719943577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,12288,0.9248417748345269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,65536,7.279448615180121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,12288,0.25816000832451713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,12288,0.8542284435696073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,65536,4.557483673095703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,10240,0.7704719967312283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,51200,3.5570657518174915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,12288,0.7646684646606445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,10240,0.7098079787360297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,8192,0.17332533995310465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,7168,0.518730640411377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,8192,0.787624888949924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,8192,0.585918214586046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,7168,0.47172090742323136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,10240,0.22590933905707467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,7168,0.15676889154646131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,6144,0.4012577798631456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,8192,0.5666791068183051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,6144,0.13702221711476645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,7168,0.5267902480231391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,6144,0.3949466546376546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,6144,0.3657040066189236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,5120,0.11656799581315781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,4096,0.253386656443278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,5120,0.2968977822197808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,4096,0.09723377890057033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,4096,0.2614933384789361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,3584,0.23630221684773764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,4096,0.2410737673441569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,51200,15.083224826388888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,3584,0.09700444671842788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,3584,0.23663645320468477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,3584,0.22155911392635771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,5120,0.33054934607611763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,2560,0.1822613345252143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,3072,0.2945244312286377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,3072,0.18991643852657744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,2048,0.14239022466871473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,2560,0.17181689209408232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,10240,0.6161680221557617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,5120,0.33056267102559406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,2048,0.058746669027540416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,1536,0.11615822050306533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,2048,0.1338248915142483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,3072,0.07713777489132352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,1024,0.07943289147482978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,1536,0.10804533296161228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,2560,0.0686968896124098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,1024,0.08580533663431804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,1024,0.041123555766211614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,768,0.061230222384134926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,1024,0.08322844240400526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,2048,0.14225867059495714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,512,0.04097244474622939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,768,0.037250667810440063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,768,0.0692159997092353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,768,0.0690586699379815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,1536,0.04820000131924947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,512,0.034448888566758894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,512,0.058641778098212353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,512,0.056872891055213086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,128,0.025976889663272437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,256,0.05261866582764519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,256,0.0448053346739875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,128,0.04948177933692932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,64,0.022834666901164587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,1536,0.1129973332087199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,32,0.023358222511079576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,64,0.05058311091528999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,256,0.031133333841959637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,2560,0.16845156086815727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,32,0.05180799961090088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,128,0.030807110998365615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,128,0.036136888795428805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,256,0.030431999100579157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,3072,0.20366399817996553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,65536,1.3924924002753363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,51200,1.0213368733723958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,16384,0.9562390645345052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,51200,2.681721793280708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,65536,3.6315875583224826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,16384,0.2833351029290093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,12288,0.6936915715535482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,16384,1.126353793674045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,16384,0.9058657752143012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,51200,4.2484283447265625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,12288,0.2165600061416626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,65536,3.8749126858181424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,12288,0.6898488998413086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,12288,0.6260968844095866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,10240,0.18517600165473092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,10240,0.5839022000630697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,8192,0.4525911013285319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,10240,0.5046302477518717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,51200,3.016966289944119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,65536,6.5846650865342875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,8192,0.44672001732720273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,7168,0.40993867980109316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,8192,0.4011502265930176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,7168,0.1325920025507609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,7168,0.3906382189856635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,6144,0.3746977647145589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,6144,0.11680355336931016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,6144,0.3437191115485297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,5120,0.31312534544203013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,8192,0.15437511603037515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,5120,0.11587288644578721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,4096,0.21270755926767984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,5120,0.2554533216688368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,4096,0.08230133189095391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,3584,0.20088888539208305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,4096,0.22212710645463732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,4096,0.2049164507124159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,3584,0.07644710938135783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,3072,0.17068000634511313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,6144,0.2959786785973443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,3584,0.18722755379147002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,5120,0.27936887741088867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,10240,0.5791501998901367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,3072,0.17075022061665854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,3072,0.16045688258277044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,2560,0.05988533629311455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,7168,0.353820456398858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,2560,0.14600622653961182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,2560,0.13676977157592773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,2048,0.04987733231650459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,3584,0.19724088244967988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,2048,0.11446489228142633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,3072,0.06540355417463514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,2560,0.14352799786461726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,1536,0.04172177778349983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,1536,0.09286222192976211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,2048,0.11535822020636664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,1024,0.03561333484119839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,768,0.05059377683533562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,1024,0.07164533270729913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,1536,0.09374222490522598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,2048,0.12079021665785049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,512,0.0373911095990075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,768,0.05944533480538262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,1536,0.09653955698013306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,512,0.05012177758746677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,256,0.031918221049838595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,1024,0.07266310850779216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,256,0.044658667511410184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,256,0.027645332945717707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,768,0.03299288948376974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,128,0.042187554968727954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,1024,0.060920000076293945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,512,0.04894399974081251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,64,0.019186667270130582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,32,0.019148444135983784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,64,0.04299199912283155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,128,0.022579555710156757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,32,0.04380177789264255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,768,0.05858044491873848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,512,0.030393779277801514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,128,0.02741777731312646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,256,0.03862933317820231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,65536,1.3570729361640081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,128,0.0318248901102278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,16384,0.7514702479044596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,51200,2.3252959781222873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,65536,3.1619894239637585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,51200,0.9321004019843208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,16384,0.2797875669267443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,16384,0.8491635322570801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,65536,3.134238137139214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,51200,3.3727849324544272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,51200,2.5157413482666016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,65536,4.70014402601454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,12288,0.49031999376085067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,12288,0.6630942026774088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,10240,0.1562026606665717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,10240,0.4976151254442003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,10240,0.4265137778388129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,8192,0.38508444362216526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,12288,0.5749342176649306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,8192,0.1288951105541653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,8192,0.3745768864949544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,7168,0.3125395509931776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,8192,0.3187546730041504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,7168,0.11221422089470758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,6144,0.2599422136942546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,16384,0.6894915368821887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,6144,0.09708088636398315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,6144,0.2730213271247016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,5120,0.21734044286939833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,6144,0.24275377061631945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,12288,0.20770755079057482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,10240,0.4535519811842177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,4096,0.1684631109237671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,5120,0.2330239878760444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,5120,0.20453243785434297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,4096,0.07020444340176053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,4096,0.18161243862575957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,3584,0.1573715607325236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,7168,0.2898168828752306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,4096,0.1703208949830797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,3584,0.06383466720581055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,3584,0.16148622830708823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,3584,0.14975644482506648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,3072,0.14043200016021729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,2560,0.11557688977983262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,3072,0.13508889410230848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,7168,0.3226835462782118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,2560,0.12043644322289361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,2560,0.11297510729895698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,3072,0.14263999462127686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,2048,0.04287999868392944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,2048,0.09498222006691827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,3072,0.05598400036493937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,5120,0.0845973359213935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,1024,0.05128533310360379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,1536,0.07975111405054729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,2560,0.049815111690097384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,2048,0.0965777768029107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,1024,0.06023999717500475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,1024,0.03038933210902744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,1024,0.060552888446384005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,2048,0.10026578108469646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,768,0.02684533264901903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,768,0.048447112242380776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,768,0.049505776829189725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,512,0.024902222885025874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,512,0.040807998842663236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,256,0.02122488949033949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,512,0.04102222124735514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,1536,0.07719288931952582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,128,0.017969777186711628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,256,0.02249777813752492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,768,0.04257066713439094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,256,0.03255555695957608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,64,0.015435554915004306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,128,0.033519999848471746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,128,0.027660444378852844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,64,0.03439733386039734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,32,0.034576889541414045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,256,0.03635288940535651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,1536,0.07253155443403456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,512,0.028316444820827905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,32,0.015523556205961438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,1536,0.035586665074030556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,128,0.022052443689770166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,65536,1.055630260043674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,65536,2.758818732367622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,51200,2.105442682902018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,51200,0.8339066505432129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,16384,0.2664719952477349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,16384,0.7093857659233941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,51200,2.9494756062825522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,16384,0.5935129059685601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,51200,2.1042026943630643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,12288,0.528796460893419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,12288,0.1635191175672743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,65536,4.47488530476888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,12288,0.5014933480156792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,10240,0.4237857659657796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,12288,0.4389982223510742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,8192,0.3295519881778293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,10240,0.46354134877522785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,8192,0.11137244436475967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,10240,0.3488062222798665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,7168,0.2891848882039388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,8192,0.28616976737976074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,16384,0.8997582329644097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,7168,0.10680355628331502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,6144,0.23962932162814668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,7168,0.24695287810431588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,6144,0.09600444634755452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,5120,0.18505777253044975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,6144,0.21911288632286918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,65536,2.6899742550320096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,8192,0.3859715461730957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,4096,0.1479520003000895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,5120,0.18367555406358507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,7168,0.28498133023579914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,4096,0.16335911220974392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,4096,0.0631635586420695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,10240,0.1391635603374905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,4096,0.15144533581203884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,6144,0.2428870995839437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,3072,0.11811199453141953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,3584,0.13589155673980713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,5120,0.07718666394551595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,3072,0.12535377343495688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,3072,0.0514008899529775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,3072,0.11923378043704563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,3584,0.14212000370025635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,3584,0.05702222055859036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,3584,0.1441297796037462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,2048,0.07621422078874376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,5120,0.21047733889685738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,2560,0.10725600189632839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,2048,0.08586666981379192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,2560,0.10762755738364325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,1536,0.07096088594860502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,2560,0.04497244291835361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,1024,0.04472177889611986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,1536,0.032724443409177996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,1536,0.0695671108033922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,1024,0.027227555712064106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,2048,0.08919644355773926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,768,0.035791999763912626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,768,0.042711112234327525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,768,0.04466933343145582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,768,0.024784889486100938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,512,0.035848001639048256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,1024,0.05362844467163086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,512,0.038154668278164335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,512,0.022133333815468684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,256,0.031046221653620403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,2560,0.10255466567145453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,256,0.029790222644805908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,128,0.016744888491100736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,128,0.02943377693494161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,128,0.02515999972820282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,128,0.020021332634819877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,64,0.016255110502243042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,32,0.016699555847379897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,64,0.029304888513353135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,512,0.025223111112912495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,32,0.02907466557290819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,2048,0.038225776619381376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,256,0.020941333638297185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,65536,0.8846506542629666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,1024,0.05400000015894572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,65536,2.033707512749566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,65536,2.383653428819444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,65536,3.611888885498047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,16384,0.5398889117770725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,51200,0.7026862038506402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,51200,1.829817877875434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,51200,1.8392694261338975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,1536,0.06395466460122003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,16384,0.21606666511959502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,51200,2.715347501966688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,12288,0.41893956396314835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,16384,0.5375582377115885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,12288,0.14646132787068686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,12288,0.42715732256571454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,12288,0.39420000712076825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,8192,0.2804995642768012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,10240,0.32128622796800405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,256,0.019113777412308585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,8192,0.30297422409057617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,8192,0.09806933005650838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,7168,0.2473599910736084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,8192,0.2558293342590332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,7168,0.08790489037831624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,6144,0.19814133644104004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,7168,0.2497048907809787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,7168,0.22724710570441353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,10240,0.38513512081570095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,6144,0.07891199986139934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,16384,0.6285075611538357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,6144,0.19609244664510092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,5120,0.06883022520277235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,5120,0.18001333872477213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,5120,0.1661235491434733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,10240,0.3480808999803331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,4096,0.05712799893485176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,10240,0.125690672132704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,3584,0.1282657782236735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,4096,0.13656889067755804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,3072,0.09806488619910346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,3584,0.051967998345692955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,5120,0.16268089082505968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,3584,0.12194577852884929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,3072,0.045448889334996544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,2560,0.08529333273569743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,3072,0.10757511191897923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,4096,0.1415475606918335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,2560,0.09462933407889472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,2560,0.09272622399859959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,6144,0.21351377169291177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,3584,0.12650933530595568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,2048,0.07865777942869398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,1536,0.05675199959013197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,2048,0.07831466860241361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,4096,0.12994666894276938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,1024,0.03479555580351088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,1536,0.06284444199668036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,1536,0.06276266442404853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,2560,0.04044888748062982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,2048,0.07630133628845215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,1024,0.04976711008283827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,768,0.03797244363360935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,2048,0.03399111164940728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,3072,0.1105511122279697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,512,0.021755556265513103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,768,0.04092266824510362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,768,0.028749333487616643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,512,0.03432710965474447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,256,0.017139555679427255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,256,0.027261333333121404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,1024,0.024159111910396155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,256,0.026891556051042344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,128,0.014534221755133735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,256,0.018142221702469718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,128,0.026047999660174053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,128,0.01792533364560869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,768,0.02186133298608992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,32,0.012747555143303342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,64,0.02557955516709222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,1024,0.04765244325002035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,32,0.02546577817863888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,512,0.02000711030430264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,64,0.012464000119103326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,65536,0.7376835611131457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,65536,1.899169709947374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,128,0.022006221943431433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,51200,1.7079537709554036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,65536,2.756355497572157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,51200,0.5834186871846517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,16384,0.5442497995164659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,1536,0.028468443287743464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,16384,0.168342219458686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,16384,0.5222462018330892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,512,0.03130666746033563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,51200,2.3341778649224176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,65536,2.0539351569281683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,12288,0.12519999345143637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,10240,0.3014150990380181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,12288,0.3713795608944363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,12288,0.3323671023050944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,10240,0.11794666449228923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,8192,0.2554328971438938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,8192,0.2391422324710422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,16384,0.47372712029351133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,8192,0.08904444509082371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,7168,0.19415911038716635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,8192,0.22330665588378906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,7168,0.07716800106896295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,7168,0.21314222282833525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,7168,0.20258755154079863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,6144,0.06979644298553467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,6144,0.18593688805898032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,6144,0.16844799783494738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,12288,0.32442845238579643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,51200,1.5797484715779622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,4096,0.09849955638249715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,5120,0.059143112765418164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,5120,0.15275466442108154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,5120,0.1432960033416748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,6144,0.16236000590854222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,4096,0.05001155535380045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,3584,0.1045866674847073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,4096,0.12192800309922959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,4096,0.11877422862582737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,3072,0.08781955639521281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,3584,0.04440711273087395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,3584,0.10812799798117743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,5120,0.151638216442532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,3072,0.03964799973699782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,3072,0.09539200199974908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,3072,0.09412178066041733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,2048,0.057532442940606006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,2560,0.08157955275641547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,2560,0.0803884466489156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,2048,0.02923111120859782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,2048,0.06746933195326063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,2048,0.0679688917265998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,3584,0.10816888676749335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,1536,0.05423466695679558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,1536,0.02516977820131514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,2560,0.03484977616204156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,1024,0.040949334700902305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,1024,0.04330044322543674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,1536,0.04912799927923414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,768,0.024296000599861145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,2560,0.0733920004632738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,768,0.019546666079097323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,512,0.018053332964579265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,512,0.02719555629624261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,10240,0.30486755900912815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,512,0.017625777257813346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,512,0.03016177813212077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,256,0.014448000325096978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,1024,0.031231111950344507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,768,0.03633244501219855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,128,0.014904000692897372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,256,0.023401778605249193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,128,0.022617777188618977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,10240,0.27441332075330943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,64,0.013839999834696451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,128,0.02030044462945726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,128,0.016806221670574613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,32,0.012672888735930124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,1536,0.0555004444387224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,64,0.022128888302379187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,256,0.016335111525323655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,1024,0.021709332863489788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,768,0.032925334241655134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,256,0.023761777414215937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,65536,1.617759068806966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,65536,2.176656934950087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,65536,1.6449982325236003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,51200,1.2328720092773438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,32,0.02239911092652215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,16384,0.37577244970533585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,51200,0.4597795804341634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,16384,0.13650933901468912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,16384,0.4101288848453098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,51200,1.5980710983276367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,51200,1.2995191150241427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,12288,0.10622221893734402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,10240,0.2359564569261339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,12288,0.275673786799113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,10240,0.24767555130852592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,65536,0.5842151112026638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,10240,0.09215555588404338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,12288,0.27128444777594674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,8192,0.20298043886820474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,16384,0.416065772374471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,7168,0.1648906601799859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,12288,0.3278782102796766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,7168,0.06506399975882636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,6144,0.14984711011250815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,7168,0.1629448864195082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,8192,0.1772151125801934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,6144,0.05786933501561483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,5120,0.10850666628943549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,6144,0.14203022585974798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,8192,0.0736008882522583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,4096,0.08842488792207505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,5120,0.050034665399127536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,5120,0.12829244136810303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,5120,0.12104177474975586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,4096,0.04134044382307265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,4096,0.1031893359290229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,4096,0.10046755605273777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,10240,0.22899201181199816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,3584,0.03698488738801744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,3584,0.09291733635796441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,3584,0.08951022227605183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,8192,0.18581067191229927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,3072,0.03242844343185425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,2560,0.05948533614476522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,7168,0.17593067222171357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,2560,0.0679253339767456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,2560,0.029098666376537744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,6144,0.15113688839806452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,3072,0.07096088594860502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,2048,0.025569778349664476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,1536,0.034040000703599714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,3072,0.08066844277911715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,2048,0.058224889967176646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,3072,0.07879110839631823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,1024,0.02574666672282749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,1536,0.02203111184967889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,1536,0.0457217792669932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,1536,0.047593779034084745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,2048,0.041695111327701144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,1024,0.018572444717089336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,1024,0.03701511025428772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,768,0.027639110883076985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,768,0.017605332864655387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,768,0.031088001198238794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,3584,0.07716444465849134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,512,0.026195555925369263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,256,0.011789333489206104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,2560,0.07004000080956353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,1024,0.033952001068327166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,256,0.014705777168273926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,256,0.019671999745898776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,128,0.01109333336353302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,512,0.015188443991872998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,128,0.014871999621391296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,128,0.01891911029815674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,64,0.009391110804345872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,128,0.017459554804695975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,32,0.009679111341635386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,512,0.01568266749382019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,768,0.019367999500698514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,32,0.018759111563364666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,2048,0.05701155794991387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,256,0.019948444432682462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,65536,0.5411138004726833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,512,0.02316711180739933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,65536,1.3332515292697482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,64,0.018609777092933655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,16384,0.34027732743157285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,51200,1.1477875179714627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,65536,1.7849884033203125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,16384,0.14458399348788792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,51200,1.3367751439412434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,12288,0.2603315512339274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,51200,1.135089768303765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,12288,0.27105511559380424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,12288,0.1086275577545166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,12288,0.25233954853481716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,65536,1.5006933212280273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,16384,0.3587964375813802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,10240,0.224634673860338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,16384,0.34520800908406574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,10240,0.20548889372083876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,8192,0.06672622097863092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,8192,0.18080355061425102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,7168,0.15475911564297146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,10240,0.22504711151123047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,7168,0.0599422189924452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,7168,0.14719377623664007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,6144,0.1368008852005005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,8192,0.16687199804517958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,6144,0.05203555689917671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,7168,0.15905688868628606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,6144,0.1309022241168552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,10240,0.08293955855899386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,5120,0.11554577615525986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,5120,0.045773333973354764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,5120,0.10979288816452026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,4096,0.09393066830105251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,4096,0.036517331997553505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,4096,0.0929866698053148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,5120,0.10818044344584148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,3584,0.03279733326699998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,6144,0.1374666690826416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,4096,0.08150844441519843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,3072,0.07247022125456068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,51200,0.421325339211358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,3584,0.08062133524152967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,2560,0.05312711000442505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,3584,0.08371377653545803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,2560,0.06093777550591362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,8192,0.17658489280276826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,2048,0.043008890416887074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,2560,0.02662044432428148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,2048,0.02292888859907786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,2048,0.051270223326153226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,1536,0.034565332863065935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,2048,0.052983111805386014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,3072,0.07224622037675646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,3072,0.02959377898110284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,1024,0.02256177696916792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,1536,0.019759999381171334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,1536,0.041300445795059204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,1536,0.04313155677583483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,768,0.01902755598227183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,1024,0.016703999704784818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,1024,0.03081688947147793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,3584,0.08332977692286174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,1024,0.03399111164940728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,2560,0.06291288799709745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,768,0.02529777752028571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,256,0.01277511070171992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,512,0.01482311056719886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,512,0.02353333267900679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,256,0.018183110488785636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,256,0.013841778039932251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,128,0.010388444695207808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,256,0.01830577850341797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,3072,0.06385066774156359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,128,0.013801777528391944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,128,0.01688799924320645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,32,0.009357333183288574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,768,0.028693334923850164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,512,0.020872889293564692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,64,0.016889777448442247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,32,0.01689511040846507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,128,0.01703644461101956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,512,0.01623288955953386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,768,0.015996444556448195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,65536,1.189684444003635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,65536,1.443036397298177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,51200,0.387847105662028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,65536,1.2890631357828777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,16384,0.2886506716410319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,64,0.009045333498054082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,16384,0.31270310613844127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,12288,0.2068479988310072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,12288,0.23655822541978624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,12288,0.22128355503082275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,10240,0.18079911337958443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,51200,0.975195566813151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,16384,0.3027555677625868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,10240,0.19701777564154732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,10240,0.0734986662864685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,8192,0.14818578296237522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,12288,0.08614399698045518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,65536,0.4694222344292535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,8192,0.059169775909847684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,16384,0.12039732933044434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,7168,0.12710755401187473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,7168,0.05334666702482435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,7168,0.1403448846605089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,6144,0.10917422506544326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,51200,0.8362320264180502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,6144,0.1210986640718248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,51200,1.1696898142496746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,8192,0.1590542263454861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,8192,0.14883644051022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,5120,0.10219377941555446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,5120,0.09850222534603542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,4096,0.06720711125267877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,6144,0.11586756176418728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,5120,0.09137955639097427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,4096,0.03273066547181871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,3584,0.0595946643087599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,4096,0.08219733503129747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,10240,0.18284444014231363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,5120,0.040052443742752075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,3584,0.02997422218322754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,3584,0.07373866770002577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,3584,0.07339021894666883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,6144,0.046781334612104625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,3072,0.026710222164789837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,3072,0.06435733371310763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,3072,0.06501511070463392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,2048,0.03607644306288825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,2560,0.05499644411934746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,2560,0.05659733215967814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,2048,0.046486222081714206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,2048,0.021244444780879553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,1536,0.028060442871517603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,2048,0.047684444321526416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,4096,0.0832186672422621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,1536,0.037296000454160906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,1536,0.03938488827811347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,2560,0.023982221881548565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,1024,0.027672888504134283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,7168,0.13569422562917074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,1024,0.030220445659425523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,3072,0.04888177911440531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,768,0.022778666681713525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,768,0.014863999353514777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,768,0.025641777449183997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,1536,0.018522666560279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,2560,0.04171644316779243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,256,0.009564444422721863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,512,0.013815999858909182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,512,0.018969777557584975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,512,0.02162844439347585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,128,0.008674666285514832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,256,0.016240888171725802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,256,0.01791466606987847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,128,0.015487111277050443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,64,0.007631999750932057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,1024,0.01938488913906945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,32,0.008923555413881937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,64,0.01521511044767168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,768,0.015565334094895257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,32,0.015218666858143277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,128,0.01312355531586541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,512,0.01238666640387641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,65536,1.0669227176242404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,65536,0.41963378588358563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,256,0.013058666553762225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,51200,0.7910239961412219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,65536,1.2848267025417752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,65536,1.1227600309583876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,16384,0.2914266586303711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,16384,0.28533244132995605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,16384,0.26665155092875165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,51200,0.8735724555121528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,1024,0.01606044504377577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,12288,0.07798666424221463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,12288,0.19832800494299996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,51200,0.9693235821194119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,10240,0.17518933614095053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,12288,0.1918640004263984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,8192,0.13710310724046496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,10240,0.1615262164009942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,12288,0.20717067188686797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,10240,0.1541973352432251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,8192,0.052689777480231396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,7168,0.10474844773610432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,8192,0.13151110543145075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,16384,0.11105066537857056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,6144,0.09671288728713989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,7168,0.12245955732133652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,7168,0.11645600530836318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,10240,0.06719999843173556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,5120,0.07794221904542711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,6144,0.04081333345837063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,6144,0.10147733158535427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,51200,0.3281795448727078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,8192,0.14172177844577366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,5120,0.03437866767247518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,5120,0.09019911289215088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,5120,0.08720178074306911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,4096,0.02899822261598375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,4096,0.07333244217766656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,4096,0.0722586645020379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,6144,0.10560266839133368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,3072,0.04968711071544223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,3584,0.06459199719958834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,4096,0.06141422192255656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,3072,0.05693511168162028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,3584,0.054564442899492055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,2560,0.04057688845528497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,3072,0.0239955551094479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,2560,0.04850133260091146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,3584,0.02629866699377696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,128,0.015959110524919298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,2048,0.03422133458985223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,2560,0.021376889612939622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,2048,0.019017777509159513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,2048,0.04082311193148295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,7168,0.04701777630382114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,2048,0.04260177744759453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,3072,0.05812355544832018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,1024,0.016546666622161865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,1536,0.016695999436908297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,1536,0.032850666178597346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,768,0.013932444983058505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,1024,0.01479466590616438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,3584,0.0653395586543613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,2560,0.051114665137396924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,1024,0.026904889278941687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,512,0.011649777491887411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,768,0.020462221569485135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,768,0.023010666171709698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,512,0.012820444173283048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,512,0.016879111528396606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,512,0.01961511042382982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,256,0.014538667268223233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,256,0.016190222567982145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,1024,0.024866667058732774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,128,0.013791110780504016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,256,0.008990222381220924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,128,0.0151582227812873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,64,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,32,0.009391999906963771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,256,0.012376888758606382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,64,0.01351288871632682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,128,0.007735999921957652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,32,0.01350933313369751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,768,0.013804444836245643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,65536,0.7073982026841906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,65536,0.33957778082953555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,51200,0.5411244498358833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,65536,1.001745753818088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,1536,0.03489777776930068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,65536,0.9024160173204211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,16384,0.1805680063035753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,51200,0.26707556512620717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,51200,0.7998817761739095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,16384,0.08479911088943481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,16384,0.23366665840148926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,128,0.012096888489193387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,10240,0.11961155467563206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,12288,0.17723822593688965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,12288,0.16596444447835287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,10240,0.14801511499616835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,10240,0.05802488989300198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,8192,0.09455022546980117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,10240,0.14055911699930826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,51200,0.676145765516493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,7168,0.08883911371231079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,8192,0.11858844757080078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,12288,0.06721422407362196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,7168,0.040419555372662015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,7168,0.10142044226328532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,12288,0.1345048877927992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,6144,0.09044622050391303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,8192,0.045653333266576133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,5120,0.06297244628270467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,6144,0.035184890031814575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,6144,0.08817866775724624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,7168,0.1044844455189175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,4096,0.04538755615552267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,5120,0.076328886879815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,4096,0.06149333053165012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,1536,0.02697511182890998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,4096,0.02544533378548092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,8192,0.11339822080400254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,3584,0.054788443777296275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,3584,0.05637244383494059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,16384,0.22184533543056914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,3584,0.02309333284695943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,3072,0.050308446089426674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,6144,0.07760266462961833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,2560,0.042075554529825844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,2048,0.017305778132544625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,3072,0.021307556165589228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,1536,0.01978755493958791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,1536,0.028778665595584448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,1536,0.030701332622104224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,1536,0.015572445260153877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,1024,0.015185778339703878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,2048,0.03513688842455546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,2048,0.024860444996092055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,1024,0.02141422198878394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,2560,0.019368888603316415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,1024,0.013458666702111563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,1024,0.023807111713621352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,2048,0.037011557155185275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,768,0.01792355544037289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,512,0.00942666663063897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,768,0.02046133412255181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,512,0.011815999945004782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,512,0.014841778410805596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,512,0.01755911111831665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,768,0.01200711147652732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,256,0.012806221842765808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,128,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,768,0.01276266657643848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,128,0.011380444798204633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,128,0.012033777932325998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,128,0.014353776971499125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,256,0.007058666812049017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,32,0.006831999868154526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,64,0.01179377817445331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,32,0.011754666765530905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,256,0.014488889111412896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,65536,0.5549448861016167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,64,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,65536,0.30169423421223956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,51200,0.4850346777174208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,65536,0.7997466723124186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,16384,0.1389360030492147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,51200,0.22596889071994355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,65536,0.7041341993543836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,51200,0.5444364547729492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,16384,0.25627289877997506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,256,0.011422221859296164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,16384,0.1846933364868164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,10240,0.09800266557269627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,12288,0.14877155092027453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,12288,0.1385013394885593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,10240,0.05538577834765116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,8192,0.07378933164808485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,10240,0.12322933144039577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,10240,0.11549066172705756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,7168,0.06649600134955512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,8192,0.0999751091003418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,16384,0.07909066809548272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,8192,0.09708711173799302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,6144,0.05549955699178907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,7168,0.03718577822049459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,12288,0.06826844480302599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,7168,0.08425955639945136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,5120,0.04649066593911913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,6144,0.03267911076545715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,6144,0.07379377550548978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,5120,0.06395022074381511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,5120,0.027564444475703772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,4096,0.03693422343995836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,12288,0.1081484423743354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,4096,0.02368533280160692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,7168,0.08797422382566665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,3584,0.030616889397303265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,4096,0.0530426667796241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,6144,0.0767493314213223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,3584,0.04696089029312134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,3584,0.047694222794638745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,8192,0.04218133290608724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,5120,0.06389688783221774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,2560,0.023549333214759827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,3072,0.04246755441029867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,4096,0.05254044466548496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,2560,0.03558222121662564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,2560,0.017283555534150865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,2560,0.03697333402103848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,3072,0.02754844393995073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,2048,0.030379556947284277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,1536,0.01573066744539473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,2048,0.031826668315463595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,3072,0.019299555155966017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,1536,0.02450399928622776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,1024,0.01179377817445331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,2048,0.019644444187482197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,3584,0.020836444364653692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,1024,0.018696889281272888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,768,0.009916444619496664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,2048,0.01626311077011956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,1024,0.02052533295419481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,3072,0.041631112496058144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,512,0.007838221887747446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,768,0.011744889120260874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,768,0.015476443701320224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,768,0.017916444275114272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,1536,0.013832888669437833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,256,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,1536,0.026009778181711834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,512,0.015651555524932016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,256,0.01037688884470198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,256,0.01110666659143236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,1024,0.012098666694429187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,51200,0.6248213450113932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,128,0.010381333529949188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,64,0.005250666704442766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,128,0.010395555860466426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,32,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,512,0.010991111397743225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,512,0.013076444466908773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,32,0.010040000081062317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,128,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,65536,0.39477867550320095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,256,0.012468444804350534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,65536,0.2191422250535753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,51200,0.30485688315497506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,65536,0.6075501971774632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,128,0.012533333566453723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,16384,0.09626755449506973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,51200,0.2003164423836602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,16384,0.15529599454667833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,16384,0.1428053379058838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,12288,0.11763644218444824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,65536,0.5512142181396484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,64,0.010187555518415239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,10240,0.06058400207095676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,12288,0.04730044470893013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,12288,0.1094355583190918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,8192,0.050125333997938365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,10240,0.09983644220564102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,10240,0.09346489111582439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,8192,0.07986044221454196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,8192,0.03118577930662367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,7168,0.043030222256978355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,8192,0.07684889104631212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,51200,0.47420533498128253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,7168,0.027387556102540758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,6144,0.03823555509249369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,7168,0.07019022438261244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,51200,0.42880622545878094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,5120,0.029535111453798082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,12288,0.07681600252787273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,6144,0.061144888401031494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,6144,0.06088178025351631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,4096,0.025265778104464214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,5120,0.051296889781951904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,5120,0.05145777927504646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,4096,0.017624000708262127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,4096,0.04297599858707852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,3584,0.02204177776972453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,4096,0.04310577776696947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,16384,0.06693333387374878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,3584,0.01679111189312405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,6144,0.024130667249361675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,3584,0.03846577803293864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,5120,0.020678222179412842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,2560,0.016720000240537856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,3072,0.03461777832772996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,2560,0.029076443778143987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,7168,0.06829066409005059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,3072,0.020032000210550096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,3584,0.03838488790724013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,2560,0.014142221874660917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,2048,0.01310488912794325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,1536,0.012057777908113269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,2048,0.025747555825445387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,1536,0.02020355562369029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,1536,0.02142133315404256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,1536,0.012014221813943652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,2048,0.014729777971903482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,2560,0.02991822361946106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,1024,0.015429332852363586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,768,0.008024888734022776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,1024,0.01723022262255351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,768,0.012857777376969656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,768,0.015170666906568738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,512,0.006695111178689533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,512,0.010926222635640038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,512,0.013144888811641268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,512,0.009886222581068674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,1024,0.009350222018029954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,256,0.005572444448868434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,1024,0.011057777537239922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,256,0.011114666859308878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,128,0.004826666580306159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,128,0.008716444174448649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,10240,0.04166222280926175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,128,0.009359111388524374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,128,0.010760000182522668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,3072,0.0337084432442983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,32,0.004559111264016894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,256,0.00942133367061615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,64,0.008354666332403818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,32,0.00833422193924586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,65536,0.2920648786756727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,768,0.010666666759385003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,65536,0.21699110666910806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,51200,0.25641687711079914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,3072,0.015250666273964776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,65536,0.5412497520446777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,16384,0.07712533076604207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,51200,0.41422579023573136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,51200,0.42826043234931094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,16384,0.13724622461530897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,16384,0.05853688716888428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,2048,0.024663110574086506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,12288,0.05695733096864489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,51200,0.17213243908352324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,12288,0.10228088829252456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,12288,0.10893333620495266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,10240,0.08720622460047404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,64,0.004263999975389904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,256,0.009356444080670675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,16384,0.14168711503346762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,8192,0.0388355553150177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,10240,0.040962666273117065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,7168,0.03471022182040744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,8192,0.030184000730514526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,8192,0.07012889120313856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,10240,0.050181332561704844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,8192,0.07608799801932441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,6144,0.029696000946892634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,7168,0.0269813338915507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,7168,0.06839199860890706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,5120,0.023582221733199224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,6144,0.05335200164053175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,6144,0.058950225512186684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,5120,0.020690666304694284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,5120,0.04517333375083076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,10240,0.0919991135597229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,5120,0.05085688829421997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,12288,0.046281778150134616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,4096,0.017654221918847825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,4096,0.03716622127426995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,4096,0.04269599914550781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,6144,0.0232968893316057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,65536,0.5217022365993923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,3584,0.038217776351504855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,4096,0.019633778267436557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,3072,0.033837334977255926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,3584,0.017270222306251526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,2560,0.025368889172871906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,3072,0.02937155630853441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,2560,0.02919644448492262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,2048,0.011639110743999481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,3072,0.015236443943447538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,7168,0.06102044714821709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,2048,0.021663111117151048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,1536,0.009696889254781935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,2048,0.025087111526065405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,3584,0.03346044487423367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,1536,0.011976888610257043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,1536,0.017906667457686532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,1536,0.021293333835071985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,2560,0.014183999763594734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,1024,0.013783111340469785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,768,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,1024,0.011077333655622272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,2048,0.013069333301650153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,3072,0.015525332755512662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,768,0.011471110913488599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,768,0.014858666393491956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,1024,0.007807999849319458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,512,0.010057777994208867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,256,0.004567111117972268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,512,0.013119999733236102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,2560,0.013742222554153867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,1024,0.017030222548378837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,256,0.00832177781396442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,256,0.009516444471147325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,256,0.010815111299355825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,768,0.010619555910428366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,64,0.003926222109132343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,128,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,32,0.003896000070704354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,64,0.007673777639865875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,32,0.007646222081449296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,65536,0.21169155173831514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,128,0.004212444441186057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,3584,0.01663822266790602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,512,0.00582755563987626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,65536,0.4404426680670844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,51200,0.17291555139753553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,128,0.01073155552148819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,16384,0.05502577622731527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,512,0.00980000032318963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,51200,0.3511937724219428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,51200,0.3086266782548693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,16384,0.046739555067486234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,12288,0.043046222792731396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,16384,0.10517866743935479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,65536,0.16807822386423746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,12288,0.08816888597276475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,10240,0.035389333963394165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,65536,0.3995564513736301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,51200,0.1341137753592597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,10240,0.03352266550064087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,10240,0.06895555390252008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,16384,0.11415466997358535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,8192,0.058916442924075656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,7168,0.02228533393806881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,8192,0.05706044700410631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,128,0.009404444032245213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,12288,0.08070311281416151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,7168,0.05194311009513008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,7168,0.02076088885466258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,6144,0.019388443893856473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,10240,0.07406666543748644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,6144,0.018352889352374606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,6144,0.04530488782458835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,5120,0.01678222252262963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,6144,0.044565333260430225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,8192,0.02347022294998169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,4096,0.013893332746293811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,5120,0.01611022154490153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,12288,0.03774133324623108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,5120,0.038246221012539335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,3584,0.012764444781674279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,4096,0.03190222382545471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,4096,0.03204177816708883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,8192,0.02658844490845998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,3072,0.011585777832402123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,3584,0.028530667225519817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,3584,0.028990222348107234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,3072,0.025263110796610515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,2560,0.010392000277837118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,3072,0.02608088817861345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,5120,0.0384897788365682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,4096,0.014197332991494073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,2560,0.021881777379247878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,2560,0.023152000374264185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,7168,0.050711999336878456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,3584,0.013458666702111563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,2048,0.011409777734014722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,2048,0.0199306673473782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,3072,0.01242844429281023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,1536,0.015562666787041558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,1024,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,1536,0.010693333215183683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,1024,0.010024000373151567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,1024,0.012123555772834353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,768,0.005591111050711737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,1024,0.013757333159446716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,1536,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,768,0.009415999882751042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,512,0.004559999952713649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,512,0.009015111459626092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,1536,0.016896888613700867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,512,0.011167999770906238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,256,0.004275555411974589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,512,0.008976889153321585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,2560,0.01203555530971951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,2048,0.00904266701804267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,128,0.0038782221575578055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,256,0.00868088917599784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,256,0.00962755580743154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,768,0.010417777631017896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,64,0.003552888830502828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,768,0.0124195557501581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,128,0.009382222261693742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,64,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,32,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,65536,0.15022133456336126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,65536,0.16608177291022405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,65536,0.3570835590362549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,51200,0.11852533287472194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,2048,0.0188044442070855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,128,0.00865955568022198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,16384,0.03619644376966689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,51200,0.13296711444854736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,51200,0.27925154897901744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,51200,0.29532623291015625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,16384,0.09422844648361206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,16384,0.10046222474839951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,256,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,12288,0.07131466600630018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,32,0.0038657776183552216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,12288,0.07742489046520658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,10240,0.05989155504438612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,16384,0.04678844412167867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,10240,0.06587821907467313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,8192,0.01845066746075948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,128,0.006713777780532837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,8192,0.054367111788855664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,7168,0.01520088811715444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,8192,0.023548444112141926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,7168,0.04288888971010844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,10240,0.03362755642996894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,7168,0.048454221751954824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,7168,0.021007110675175984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,65536,0.3763057655758328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,12288,0.037801778978771634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,6144,0.042429334587521024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,5120,0.03209422363175286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,6144,0.012110222544935016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,5120,0.01626844373014238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,5120,0.03624177641338772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,10240,0.023617777559492324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,5120,0.011218667030334473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,4096,0.02641422218746609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,4096,0.014491554763582019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,3584,0.008793777889675563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,4096,0.030560887522167627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,3584,0.023614222804705303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,3584,0.027456889549891155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,8192,0.04905244376924303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,3072,0.007633777956167857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,6144,0.03748177819781833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,6144,0.018445332845052082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,3072,0.012732444538010491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,3072,0.021060445242457922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,3072,0.025073778298166063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,12288,0.0284204449918535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,2048,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,2560,0.012096000214417776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,2560,0.02198133369286855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,2048,0.011507555842399597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,4096,0.009036444127559662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,1536,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,2048,0.018914666440751817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,1536,0.01071466671095954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,1536,0.013199110825856527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,1536,0.016210666961140104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,2560,0.018410666121376883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,1024,0.010586666564146677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,768,0.004311999926964442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,1024,0.009923555784755284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,3584,0.01349866638580958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,512,0.003912888881233004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,768,0.009047110875447592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,768,0.011738667057620155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,2560,0.007222221957312689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,1024,0.004554666578769684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,512,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,512,0.010777778095669217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,256,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,128,0.0035626664757728577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,256,0.00907555553648207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,2048,0.015849777393870883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,768,0.00941777808798684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,64,0.003236444460021125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,128,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,32,0.0032008888406885993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,256,0.003540444291300244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,64,0.005643555687533484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,65536,0.1018399993578593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,32,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,256,0.008667555948098501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,1024,0.013455111119482251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,51200,0.07934133211771648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,65536,0.3168879879845513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,128,0.00869688888390859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,128,0.00906844437122345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,16384,0.03122666809293959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,51200,0.25428978602091473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,512,0.00793422261873881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,51200,0.29106222258673775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,16384,0.04780266682306925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,16384,0.08303733004464044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,16384,0.09894133276409572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,10240,0.016700444949997794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,12288,0.03849333193567064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,12288,0.07643200291527642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,65536,0.3704826566908095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,10240,0.034268445438808866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,8192,0.014170666535695394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,10240,0.0649813347392612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,12288,0.02202755543920729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,7168,0.012099555797047086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,8192,0.05384711093372769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,12288,0.06298133399751452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,65536,0.1654791169696384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,7168,0.0383857786655426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,7168,0.02109777761830224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,7168,0.047621332936816745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,10240,0.05334488881958855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,5120,0.009365333451165093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,6144,0.03342933456103007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,6144,0.0421324438518948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,8192,0.023687111006842718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,8192,0.04391111267937554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,5120,0.028370665179358587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,5120,0.01702666613790724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,5120,0.03536533315976461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,3584,0.007654222349325816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,4096,0.014863999353514777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,4096,0.023540443844265405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,4096,0.029851555824279785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,3072,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,51200,0.13718756039937338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,3584,0.02714311083157857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,3072,0.019090667366981506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,2560,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,3072,0.02408711115519206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,2560,0.016883555385801528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,6144,0.01071999967098236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,2048,0.005620444400442972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,2560,0.021597334080272253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,3584,0.0140337778462304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,6144,0.019300444258583915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,1536,0.004886222382386525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,3072,0.01315111087428199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,4096,0.008040889269775814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,1536,0.012182222472296821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,1536,0.010703110860453712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,2560,0.012318222059143914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,1024,0.0042168887125121225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,2048,0.011743110915025076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,2048,0.014541332920392355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,1024,0.009776888622177972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,768,0.004009777886999978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,1024,0.01309333327743742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,1024,0.009907555249002244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,512,0.0038764443662431505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,768,0.009576000273227692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,768,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,768,0.011767999993430244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,256,0.003527111063400904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,1536,0.016181333197487723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,512,0.010372444159454769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,256,0.008671999805503422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,128,0.0029226665695508323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,256,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,2048,0.018574222922325134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,128,0.008656000097592672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,128,0.005591111050711737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,64,0.0028417776856157514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,32,0.0031999999450312722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,512,0.009045333498054082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,64,0.00526577772365676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,65536,0.09071999788284302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,32,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,51200,0.07438577546013726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,256,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,16384,0.030810667408837214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,12288,0.021495999561415777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,10240,0.019157333506478202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,65536,0.32596445083618164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,51200,0.2467715475294325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,16384,0.08308977550930448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,7168,0.011749332977665795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,10240,0.05315466721852621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,512,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,6144,0.010407110883129967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,8192,0.04315911067856682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,5120,0.009343110852771336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,3584,0.021358221769332886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,4096,0.008016889293988546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,7168,0.03808355662557814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,3584,0.007342221836249034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,3072,0.006671111202902264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,5120,0.028171555863486394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,4096,0.0237128883600235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,3584,0.02131822208563487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,2048,0.005586666779385672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,12288,0.06294933292600843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,1536,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,2048,0.014674666855070325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,1024,0.004232888834344016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,1536,0.012103111379676394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,768,0.003887999802827835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,6144,0.03291555576854282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,512,0.003567999849716822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,1024,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,256,0.0031964443624019623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,768,0.008395555946562026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,128,0.0028817777832349143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,512,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,2560,0.016799110505315993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,256,0.005924444645643234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,32,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,128,0.0052471111218134565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,65536,0.09595111343595718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,32,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,2560,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,51200,0.07639644543329875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,8192,0.01349422252840466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,65536,0.31530578931172687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,128,0.008416888614495596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,12288,0.020676443974177044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,10240,0.012436444560686747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,16384,0.08453866508271958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,64,0.005250666704442766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,12288,0.06279200315475464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,7168,0.010034666293197209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,8192,0.04308977723121643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,6144,0.009377777576446533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,16384,0.02793866727087233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,5120,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,64,0.0029084444459941653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,4096,0.008008889026112026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,6144,0.03302844365437826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,3584,0.007666666474607255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,5120,0.028007109959920246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,8192,0.010751999914646149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,3072,0.01240533341964086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,2560,0.01072533345884747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,10240,0.05303022265434265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,2048,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,3072,0.019553777244355943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,1536,0.004590222405062782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,3072,0.018954666124449838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,7168,0.037900444534089826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,2048,0.014597333139843412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,1024,0.003921777837806278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,768,0.003928889003064898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,1536,0.012319111161761813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,512,0.003532444437344869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,1024,0.009767110976907942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,4096,0.023289778166347083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,768,0.008513777620262569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,128,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,512,0.007284444239404466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,64,0.0028524444335036804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,256,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,32,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,128,0.0053075556125905775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,64,0.005246222019195557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,51200,0.2467528978983561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,256,0.0031964443624019623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,3584,0.021948445174429152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,2560,0.01715022159947289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,32,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,16384,1.8582168155246312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,12288,1.2695973714192708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,12288,4.353068457709418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,16384,5.194441901312934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,12288,4.236328125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,16384,6.043497721354167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,10240,1.12737062242296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,51200,7.307020399305556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,8192,0.8484773106045194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,8192,2.86481073167589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,10240,3.3536302778455944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,8192,2.679595523410373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,7168,0.7400373352898492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,10240,3.7545255025227866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,6144,0.6711840099758573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,7168,2.7149236467149525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,6144,2.3825013902452254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,6144,1.9015492333306208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,5120,0.5460737546284994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,7168,2.3440897199842667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,4096,1.290276421440972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,5120,1.6467537350124781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,4096,0.4501484235127767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,5120,1.6923866271972656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,4096,1.2755813598632812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,3584,0.4116560088263617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,3584,1.0791039996676974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,3072,0.36690666940477157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,3072,1.0849475860595703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,3584,1.1068390740288627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,3072,0.9545528623792859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,2560,0.31704621844821507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,2560,0.9339484108818902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,2048,0.74890046649509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,2048,0.266612450281779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,1536,0.49311288197835285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,2048,0.6604497697618272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,1536,0.5326773325602213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,1536,0.24441244867112902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,1024,0.37810667355855304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,1024,0.16670933034684923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,768,0.1577262216144138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,512,0.21723911497328016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,768,0.34382134013705784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,2560,0.8181537522210015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,256,0.13780266708797878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,512,0.2629937860700819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,256,0.11345422267913818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,256,0.2373502254486084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,768,0.26149243778652614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,128,0.11339467101626927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,64,0.10484444432788426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,1024,0.3921182155609131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,128,0.2232337792714437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,512,0.13019643889533147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,64,0.23114222950405547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,32,0.23576890097724068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,51200,17.186321682400173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,32,0.11096711291207208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,128,0.12012444602118598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,16384,4.229744805230035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,51200,5.474190182156033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,51200,25.728418986002605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,65536,7.081758287217881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,16384,1.3939652972751195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,12288,3.112566206190321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,12288,1.075914700826009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,16384,4.8098326789008246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,12288,3.5595164828830295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,10240,0.8184231122334799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,51200,13.237452189127604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,10240,2.8287181854248047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,10240,2.7218543158637156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,8192,2.4281510247124567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,8192,2.1126151614718967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,8192,0.6688577863905164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,7168,0.5900390942891439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,7168,1.9060906304253473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,7168,1.7994115617540147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,65536,16.47841559516059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,6144,0.5229679743448893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,6144,1.5424310896131728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,5120,0.43849778175354004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,5120,1.2344719568888347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,6144,1.6660772959391277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,4096,1.0981875525580513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,5120,1.2603848775227864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,4096,1.0265369415283203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,3584,0.7918408711751302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,3584,0.32244443893432617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,3584,0.8924658033582898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,4096,0.36852889590793186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,3072,0.27957243389553493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,3072,0.7926230960422092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,2560,0.24834489822387695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,2560,0.5953546630011665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,3072,0.768487983279758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,2048,0.5287093586391872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,51200,19.473559909396702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,2048,0.208624005317688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,1536,0.4351217746734619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,1536,0.18900088469187418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,1024,0.2886426713731554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,1536,0.41100621223449707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,768,0.22144444783528647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,1024,0.3084577719370524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,512,0.15134578280978733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,768,0.11645600530836318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,768,0.28547289636400014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,256,0.12278489271799724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,512,0.2088408867518107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,2560,0.631391101413303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,256,0.0907519989543491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,2048,0.562089761098226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,256,0.18891644477844238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,64,0.08242933617697822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,128,0.09119288788901435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,1024,0.15109422471788195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,64,0.18289600478278267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,32,0.186027553346422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,512,0.10880800088246663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,128,0.09472622474034627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,32,0.08751822180218166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,65536,24.418431599934895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,128,0.1756124496459961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,65536,2.198058658175998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,51200,1.688344849480523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,16384,0.4565475781758626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,16384,1.5654364691840277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,16384,1.4636995527479384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,12288,1.1098480224609375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,51200,4.207665761311849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,12288,0.32201600074768066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,10240,0.27354489432440865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,10240,0.7786400053236219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,8192,0.6432017750210232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,8192,0.23736977577209473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,10240,0.9040213690863715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,7168,0.20477422078450522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,8192,0.6949449115329318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,65536,5.2482727898491754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,12288,1.0641875796847873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,7168,0.6012861993577745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,6144,0.5170373386806911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,6144,0.17890577846103242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,5120,0.4117697874704997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,6144,0.5035431120130751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,5120,0.1572835577858819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,4096,0.33046666781107586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,7168,0.621196428934733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,3584,0.2833920054965549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,51200,7.238833957248264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,4096,0.35255999035305446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,3584,0.10946400298012628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,3072,0.2602906756930881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,65536,8.142390780978733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,2560,0.20542755391862658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,3072,0.2835120095147027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,2560,0.09134222401512994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,2048,0.16652622487809923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,2560,0.23491289880540636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,1536,0.12228088908725315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,2048,0.07188622156778972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,4096,0.1229706605275472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,2048,0.18064267105526396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,1536,0.05905955367618137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,1536,0.14272444778018528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,1024,0.0487937761677636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,1024,0.10467733277214898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,768,0.044657776753107704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,768,0.08423378070195515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,512,0.04092266824510362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,512,0.07157422436608209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,256,0.03981155488226149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,1024,0.09106933408313328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,128,0.03211288981967502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,256,0.0644542243745592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,3072,0.09727822409735785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,128,0.03633688886960348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,512,0.053320888015958995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,128,0.060510224766201444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,32,0.030096001095241968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,5120,0.4178275532192654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,32,0.06532355811860827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,3584,0.29962044292026097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,64,0.028470221492979262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,64,0.06246933009889391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,256,0.03672711054484049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,65536,1.6180640326605902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,16384,1.1184284422132704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,51200,1.2714204788208008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,768,0.07304977708392672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,16384,0.36957510312398273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,12288,0.2595253255632189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,12288,0.7838586701287164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,65536,4.246706644694011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,16384,1.141418668958876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,10240,0.22157511446211073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,10240,0.7448204358418783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,8192,0.5206613540649414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,10240,0.6474791102939182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,12288,0.8544044494628906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,7168,0.4669031037224664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,7168,0.15412533283233643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,7168,0.4413013458251953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,6144,0.4506533410814073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,6144,0.13357422086927626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,5120,0.3411244551340739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,51200,3.182859632703993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,5120,0.13804354932573107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,4096,0.254384888543023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,5120,0.3157715532514784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,51200,4.678782992892795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,4096,0.0941022237141927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,4096,0.2572462293836805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,65536,6.060288747151692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,3584,0.08454755279752944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,3072,0.19325866964128283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,8192,0.2430337799919976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,2560,0.16118844350179037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,3072,0.19166933165656197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,8192,0.5211111174689399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,2560,0.1640684472190009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,3584,0.21873066160413954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,2048,0.05727022224002414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,2048,0.13576000266604954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,1536,0.11109333568149143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,6144,0.38330401314629453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,1024,0.07554399967193604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,1536,0.04716444346639845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,3584,0.22301332155863443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,1536,0.10792711046006943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,1024,0.039693332380718656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,1024,0.08112888866000705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,768,0.036335110664367676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,768,0.0655591090520223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,2048,0.1278266641828749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,512,0.03353688783115811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,512,0.05566133393181694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,256,0.03024266825781928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,256,0.05003289050526089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,128,0.030051555898454454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,128,0.0469466679626041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,64,0.02224622170130412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,32,0.023380445109473333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,2560,0.06786755720774333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,64,0.04832888974083794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,32,0.04933511217435201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,128,0.02509511179394192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,3072,0.07506133450402154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,512,0.04435466726620992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,256,0.030403554439544678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,65536,1.460995568169488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,768,0.06629333231184217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,51200,1.0735875235663521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,16384,0.29286932945251465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,65536,3.3911955091688366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,12288,0.7190924220614963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,51200,2.9923608567979603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,16384,0.7730035781860352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,12288,0.6600160068935818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,10240,0.18164266480339897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,10240,0.5710542466905382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,8192,0.454544915093316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,10240,0.6995902061462402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,8192,0.4348071151309543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,7168,0.37503555085923934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,16384,0.9801493750678169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,12288,0.2198559972974989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,51200,4.233717176649305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,7168,0.38070045577155215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,6144,0.11665866110059951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,8192,0.21726932790544298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,65536,6.21997790866428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,7168,0.13014666239420572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,6144,0.3335786660512288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,4096,0.08038489023844402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,5120,0.2651991049448649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,3584,0.07268889082802667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,3584,0.18874755170610216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,4096,0.2109760046005249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,3072,0.1580293311013116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,3072,0.06385244263543023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,3072,0.16269421577453613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,2560,0.13313778241475424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,2560,0.056341330210367836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,2560,0.1391786601808336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,2048,0.10898666911655003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,2048,0.048376888036727905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,2048,0.11602132850223118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,4096,0.24050132433573404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,6144,0.3198311063978407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,1024,0.06441955433951484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,5120,0.10021422306696574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,1024,0.03455911080042521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,1024,0.0699911117553711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,768,0.03233155608177185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,768,0.05603377686606514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,3584,0.20842133627997506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,1536,0.09250311056772868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,512,0.029663112428453233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,768,0.05010399884647793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,256,0.04257955484920078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,128,0.021560000048743352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,512,0.03447022371821933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,128,0.026535110341178045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,1536,0.04064622190263536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,512,0.04787733488612705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,32,0.020012444920010038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,64,0.04120799899101257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,256,0.02678044471475813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,128,0.04018044471740723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,64,0.019271110494931538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,256,0.02693333393997616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,32,0.0421013335386912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,5120,0.27026134067111546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,65536,2.8621927897135415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,1536,0.08817778031031291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,16384,0.6963493559095594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,51200,0.9589049021402994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,16384,0.260171545876397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,65536,1.258401764763726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,12288,0.5608008702596029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,12288,0.1855306625366211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,16384,0.7890124320983887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,10240,0.4675644238789876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,12288,0.6226142247517904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,10240,0.1545119947857327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,8192,0.3719173272450765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,8192,0.12900977664523655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,10240,0.5867377916971842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,7168,0.2991680039299859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,65536,4.558678944905599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,8192,0.391795555750529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,6144,0.26774488555060494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,6144,0.10448177655537923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,7168,0.34114755524529355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,5120,0.23763555950588652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,6144,0.2651386790805393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,4096,0.15586222542656794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,51200,2.0634382035997176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,5120,0.23806844817267525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,4096,0.17433155907524955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,3584,0.062181333700815834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,3072,0.11811910735236274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,3584,0.15380177232954237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,7168,0.1122337712181939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,3072,0.05682933330535889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,3072,0.13560088475545248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,2048,0.0856319997045729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,5120,0.08337066570917766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,4096,0.0692302253511217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,1536,0.06579733557171293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,2048,0.09587022331025864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,1536,0.03491377830505371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,1536,0.07605599694781832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,2560,0.09688444270028009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,2560,0.04917600088649326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,1024,0.05726222197214762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,3584,0.13683556185828313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,2048,0.041305777099397444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,768,0.02618666655487484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,512,0.027215111586782668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,512,0.02368444369898902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,512,0.03854577740033468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,1024,0.029267556137508813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,768,0.038402666648228966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,128,0.017459554804695975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,256,0.02117066747612423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,128,0.02095199955834283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,64,0.015513777732849121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,1024,0.045704887972937695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,32,0.01588088936275906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,256,0.02090488870938619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,64,0.032038221756617226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,2560,0.11460444662306045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,32,0.03199999862247043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,256,0.033816887272728816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,768,0.04593866732385424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,128,0.03183555603027344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,51200,3.26325692070855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,65536,1.077888912624783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,51200,0.8467644585503472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,65536,2.26490232679579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,16384,0.6210826767815484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,16384,0.2794728808932834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,51200,2.2953368292914496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,12288,0.47524176703559023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,16384,0.8059884707132975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,12288,0.177419556511773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,10240,0.410785781012641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,10240,0.14181599352094862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,12288,0.635457780626085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,8192,0.3451208803388808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,65536,3.511072794596354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,7168,0.2786213292015923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,8192,0.3704293304019504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,6144,0.2233955595228407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,7168,0.2946346600850423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,5120,0.17052177588144937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,6144,0.23180267545912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,4096,0.150435553656684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,5120,0.1949982245763143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,8192,0.11445244153340657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,51200,3.2471351623535156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,4096,0.15460177262624106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,7168,0.09893600145975749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,3584,0.0571075545416938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,3584,0.13771022690667048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,3072,0.05029866761631436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,3072,0.1197928852505154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,2560,0.10261599885092841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,10240,0.48052178488837344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,3584,0.13641244835323757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,2048,0.08590044577916463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,3072,0.1153235567940606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,2048,0.03733866744571262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,6144,0.08631377749972874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,1536,0.059939556651645236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,2048,0.08521599902047051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,1024,0.04064177804523044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,1536,0.03127466638882955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,768,0.032116442918777466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,1024,0.02623555560906728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,1024,0.050910221205817334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,512,0.02409155501259698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,768,0.023599111371570166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,768,0.04123199979464213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,2560,0.1028675569428338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,512,0.021464889248212177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,512,0.03394577900568644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,256,0.019136000010702345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,256,0.029480000336964924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,5120,0.07435733079910278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,128,0.019256888164414298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,64,0.013487111363146039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,32,0.014111111561457316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,4096,0.06364888615078397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,64,0.027828445037206013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,256,0.019302222463819716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,32,0.027787556250890095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,128,0.015963556038008798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,1536,0.06799288590749104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,128,0.028060442871517603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,2560,0.044365333186255566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,65536,0.8985297944810656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,16384,0.4931813346015082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,51200,1.5248364342583551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,16384,0.24146933025783965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,51200,0.6782053311665853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,12288,0.36237332555982804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,16384,0.5760382016499838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,12288,0.14896889527638754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,65536,2.30602052476671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,10240,0.4347502125634088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,10240,0.12384177578820123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,12288,0.4375511010487874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,10240,0.3390177885691325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,7168,0.2512871159447564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,8192,0.2870817714267307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,7168,0.08644711309009129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,51200,2.545045428805881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,6144,0.2074835565355089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,7168,0.24016888936360678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,6144,0.07617955737643771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,5120,0.06564978096220228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,5120,0.171470218234592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,8192,0.2692844337887234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,4096,0.05612444215350681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,3584,0.1222497754626804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,3584,0.050256000624762646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,65536,3.1153990427652993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,5120,0.16828355524275038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,6144,0.20554044511583117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,3072,0.0440906650490231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,2560,0.08656977944903904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,3072,0.10553066598044501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,2048,0.06364711125691731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,4096,0.13695911566416422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,2560,0.09022844500011867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,2048,0.032808889945348106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,3584,0.1212631066640218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,1536,0.05060977737108866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,4096,0.12685867150624594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,1024,0.035790221558676824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,2560,0.03871999846564399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,1024,0.04515999886724684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,1024,0.023620444867346022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,768,0.02645422187116411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,2048,0.07534044318728976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,3072,0.10643555720647176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,768,0.02109066645304362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,768,0.03663644525739882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,256,0.017116443978415597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,1536,0.027787556250890095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,1536,0.059975114133622914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,256,0.017535999417304993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,256,0.026176000634829204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,8192,0.11763111750284831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,64,0.0148053334818946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,128,0.024844444460339014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,32,0.015823110938072205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,64,0.024734222226672705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,512,0.024312888582547505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,512,0.01902844508488973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,512,0.02995377779006958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,128,0.017595556047227647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,65536,0.6967013147142199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,32,0.024868445263968572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,51200,1.2369271384345162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,16384,0.4369084570142958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,65536,2.0983439551459417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,128,0.013760888742076026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,12288,0.12592889202965632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,16384,0.4904097980923123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,12288,0.39426043298509383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,10240,0.3001084327697754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,51200,2.046529769897461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,65536,2.7959456973605685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,10240,0.10820444424947102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,51200,0.5664186477661133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,10240,0.29166311687893337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,8192,0.09250489208433364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,8192,0.24704445732964408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,7168,0.07753866910934448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,6144,0.16016711129082575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,16384,0.17784089512295195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,7168,0.20363910992940268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,5120,0.1279484430948893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,6144,0.17595822281307646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,5120,0.05783200263977051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,5120,0.145505772696601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,8192,0.21952177418602836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,3584,0.08997066815694173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,4096,0.11801244152916802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,7168,0.20677244663238525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,3072,0.08736799822913276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,3584,0.10357866684595744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,12288,0.34058133761088055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,3072,0.03859466645452712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,6144,0.06755110952589247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,2048,0.055889778667026095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,2560,0.07813777526219685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,4096,0.047988444566726685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,1536,0.04477689001295301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,3584,0.04335822330580818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,1536,0.02440533373090956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,1024,0.027523555689387854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,3072,0.09059821897082859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,2560,0.03403910994529724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,768,0.023181334137916565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,1024,0.03933688998222351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,2048,0.06499110990100436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,2048,0.029291556941138372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,512,0.01755555636352963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,512,0.017106667160987854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,256,0.013728000223636627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,4096,0.10591910945044623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,1024,0.021074665917290583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,128,0.012063110868136087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,768,0.018912000788582694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,768,0.03170488940344916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,64,0.01074488874938753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,128,0.016043555405404832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,32,0.01074222226937612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,2560,0.06150044335259331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,64,0.021360889077186584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,32,0.021347555849287245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,256,0.022337777747048274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,256,0.015970667203267414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,128,0.021623111433453027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,65536,0.5851653416951498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,1536,0.05191555619239807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,65536,1.4652844534979925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,16384,0.3541902171240912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,51200,1.209103160434299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,16384,0.15940799978044298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,16384,0.39274756113688153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,12288,0.10813777976565891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,12288,0.40331289503309464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,512,0.025928889711697895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,10240,0.08852177858352661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,51200,0.4755031267801921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,10240,0.2425626648796929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,12288,0.2715848816765679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,8192,0.07221600082185534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,7168,0.16138488716549343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,65536,2.034121831258138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,6144,0.12347022692362468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,7168,0.16889688703748915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,6144,0.05666489071316189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,6144,0.1458337836795383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,8192,0.18506488535139295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,5120,0.11060177617602879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,4096,0.08415555291705662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,5120,0.04959911108016968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,51200,1.5484790802001953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,5120,0.12331110901302761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,3584,0.07196799914042155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,7168,0.06355644596947564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,4096,0.0996862252553304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,3072,0.07248622179031372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,3584,0.08880177471372817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,3072,0.07675377527872722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,2560,0.06543555524614122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,2048,0.04217866725391812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,10240,0.18875466452704537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,8192,0.1901857720481025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,4096,0.040489776266945735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,2048,0.024953777591387432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,2048,0.054733332660463124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,3584,0.035862220658196345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,1024,0.02437866727511088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,2560,0.05298755566279093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,1536,0.044216887818442456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,1024,0.018225777480337355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,1024,0.03264177838961283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,768,0.016535111599498324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,2560,0.028868443436092798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,768,0.026808889375792608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,256,0.011516444385051727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,512,0.015236443943447538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,1536,0.038062221474117704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,512,0.02203111184967889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,128,0.010072000324726105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,256,0.01444533301724328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,256,0.019144889381196763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,1536,0.021681777305073206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,3072,0.032080888748168945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,128,0.01826844447188907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,64,0.017886221408843994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,32,0.01791733337773217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,64,0.00904888908068339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,128,0.014476444986131458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,768,0.019848000672128465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,65536,1.3404426574707031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,512,0.014736889137162102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,16384,0.36888355678982204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,51200,0.42305776807996964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,51200,0.9126817915174695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,65536,1.662975099351671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,16384,0.15264444881015354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,12288,0.24314488304985893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,16384,0.34191732936435276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,12288,0.09774844513999091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,10240,0.2008986605538262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,32,0.009015999734401703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,12288,0.2515653239356147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,10240,0.08157244655821058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,7168,0.13453777631123862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,10240,0.22008355458577475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,8192,0.1763146718343099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,6144,0.12509689066145155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,7168,0.15157955222659641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,5120,0.09470844268798828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,6144,0.13130133681827122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,65536,0.5240497589111328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,5120,0.04376533296373156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,4096,0.08082044124603271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,51200,1.360489739312066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,3584,0.07557777563730876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,4096,0.03577422102292379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,4096,0.09062311384412979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,7168,0.058196445306142174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,3072,0.06285155481762357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,3584,0.0793093310462104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,6144,0.05117155445946587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,2560,0.050181332561704844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,3072,0.06906311379538642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,2560,0.026259554757012263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,2560,0.05832622448603312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,8192,0.06502488586637709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,2048,0.04925688770082262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,3584,0.03235022226969401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,1536,0.03936800029542711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,3072,0.029382222228580054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,1024,0.021822222405009802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,2048,0.03993599944644504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,1024,0.016399110356966656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,768,0.01791733337773217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,1024,0.02938577863905165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,2048,0.022289777795473736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,1536,0.028348445892333984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,768,0.02427644365363651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,512,0.014494222071435718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,512,0.020047111643685233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,256,0.017301332619455125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,256,0.013624889155228933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,512,0.014045332868893942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,128,0.016568000117937725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,128,0.013502221968438892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,64,0.008380444513426887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,5120,0.11054666837056477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,32,0.00869066682126787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,768,0.015787555111779105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,128,0.009399111072222391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,32,0.01629244453377194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,256,0.010759111079904767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,65536,0.4472168816460504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,8192,0.16281067000495061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,65536,1.1290826797485352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,51200,0.8536551263597277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,16384,0.2620168791876899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,51200,0.38753777080112034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,16384,0.1305102242363824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,12288,0.19421866205003527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,16384,0.3143724335564507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,51200,1.116634686787923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,10240,0.17130222585466173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,12288,0.23294311099582246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,8192,0.13927822642856175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,10240,0.07863377862506442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,10240,0.19581511285569933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,8192,0.05807555384106106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,8192,0.15259377161661783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,64,0.016375111209021676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,7168,0.0527724458111657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,7168,0.1339893341064453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,1536,0.019324445062213473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,6144,0.11510845025380452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,12288,0.0906791090965271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,4096,0.0640888876385159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,5120,0.03829866647720337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,5120,0.09875910811954075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,4096,0.03212177753448486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,3584,0.0649048884709676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,3584,0.029044446018007066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,6144,0.10278577937020196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,3072,0.05110044280687968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,5120,0.08626221948199803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,3072,0.026296888788541157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,2560,0.045695109499825366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,3072,0.06156000163820055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,2048,0.03686044282383389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,2560,0.023744889431529578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,2560,0.053005334403779775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,2048,0.020850666695170935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,4096,0.07995288901858859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,2048,0.0444124440352122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,1024,0.020033778415785897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,1536,0.018220444520314533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,1536,0.03565244542227851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,768,0.015547555353906421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,6144,0.04571555720435249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,1024,0.026534222894244727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,512,0.01218311074707243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,768,0.014720888601409065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,768,0.021830222672886316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,256,0.009726222190592024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,7168,0.12897777557373047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,512,0.018283555905024212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,128,0.00830666638082928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,256,0.015850666496488783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,1024,0.01573244399494595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,64,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,512,0.013471110827393003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,64,0.014828445182906257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,32,0.007640000018808577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,256,0.012814222110642327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,65536,1.501484447055393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,32,0.014831999937693277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,128,0.012776888906955719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,65536,0.8355786535474988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,51200,0.7171840137905545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,51200,0.33718665440877277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,3584,0.07021155622270372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,51200,0.8861582014295789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,128,0.014890667464998035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,16384,0.10029066933525933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,12288,0.17726488908131918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,16384,0.26104532347785103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,65536,0.4003013239966498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,12288,0.19732533560858834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,1536,0.02868355645073785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,8192,0.11156177520751953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,10240,0.16431111759609646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,65536,1.1657821867201064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,8192,0.05121689041455587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,8192,0.1342453294330173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,7168,0.045977777904934354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,7168,0.11643821663326687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,10240,0.15326578087276882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,6144,0.03926577832963731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,6144,0.10088088777330186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,5120,0.07671199904547797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,16384,0.2428969012366401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,7168,0.09699289004007976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,5120,0.08607911401324803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,6144,0.08389777607387966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,4096,0.06929333342446221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,3584,0.04573422339227465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,3584,0.02589688864019182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,3072,0.03858133488231235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,4096,0.057292441527048744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,3072,0.05348799957169426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,3072,0.023544889357354905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,4096,0.029030223687489826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,12288,0.07780533366733126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,2048,0.02905244297451443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,2560,0.046331554651260376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,10240,0.06511377625995211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,3584,0.061087999078962535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,2048,0.0389066669676039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,1536,0.016255999604860943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,1536,0.031204445494545832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,2560,0.03435111045837402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,768,0.01254488858911726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,2560,0.02132622235351139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,768,0.019533332851197984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,768,0.013776000175211163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,1024,0.015689777003394235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,512,0.016565332810084026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,512,0.01275022245115704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,256,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,1024,0.023602666126357183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,256,0.012268444730175866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,128,0.007601777712504069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,256,0.014185777968830533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,64,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,1536,0.021164443757798936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,32,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,5120,0.0339626669883728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,64,0.013162666724787818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,1024,0.014551111393504672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,65536,0.6929670969645182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,128,0.01204533295498954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,2048,0.01847555571132236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,65536,0.9401271608140734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,51200,0.5793529086642796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,51200,0.25990221235487193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,32,0.013117333253224691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,16384,0.08271377616458468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,51200,0.7412897745768229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,16384,0.26284533076816136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,65536,0.3310328854454888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,10240,0.11603911717732747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,128,0.013447999954223633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,12288,0.16903200414445665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,8192,0.08769955237706502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,10240,0.05724622143639458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,10240,0.14118756188286677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,16384,0.18368177943759492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,8192,0.11371466848585342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,7168,0.03902666767438253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,7168,0.10017689069112141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,12288,0.06491021977530585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,512,0.010021333065297868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,5120,0.0539244446489546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,5120,0.03000088863902622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,7168,0.08678578005896674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,5120,0.07403733332951863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,4096,0.02509333358870612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,6144,0.0644035538037618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,3584,0.0369102226363288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,6144,0.03401599989997016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,3584,0.05323466658592224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,3072,0.03228088882234361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,3072,0.0210853334930208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,8192,0.04394666684998406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,12288,0.12909066677093506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,2560,0.040806220637427434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,2048,0.02456355591615041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,2560,0.0188755558596717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,3584,0.022945778237448797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,6144,0.08692089054319595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,2048,0.016920000314712524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,2048,0.03451733456717597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,1024,0.014167999227841696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,1536,0.015132443772421943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,4096,0.04183200001716614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,768,0.011735999749766456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,1024,0.013454222016864352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,1024,0.02093066606256697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,512,0.009379555781682333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,768,0.01721155477894677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,512,0.011757333245542316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,1536,0.018619555566045973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,256,0.0069520001610120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,256,0.011078221930397881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,3072,0.046925332811143666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,128,0.006269333263238271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,768,0.012416000167528788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,128,0.011765333513418833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,64,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,512,0.014534221755133735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,32,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,64,0.011704000333944956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,1536,0.02755644420782725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,32,0.01145600030819575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,65536,0.5281662411159939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,65536,0.26961689525180393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,128,0.011056889262464313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,51200,0.46512354744805234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,65536,0.7380506727430555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,51200,0.2128222253587511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,16384,0.14800177680121526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,2560,0.02972266740269131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,16384,0.07199911276499431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,51200,0.5894826783074273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,16384,0.1832782162560357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,10240,0.09322222073872884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,12288,0.061518225404951304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,12288,0.13981866836547852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,8192,0.07621333334181044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,10240,0.05394488904211256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,10240,0.11664977338578965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,8192,0.04114755656984117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,8192,0.09490133656395806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,7168,0.035573333501815796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,7168,0.08313688966963026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,256,0.012430222498046027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,5120,0.043875555197397866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,6144,0.07215822405285306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,4096,0.03569777806599935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,4096,0.05932888719770643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,5120,0.060531556606292725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,3584,0.032136887311935425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,4096,0.022896000080638464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,6144,0.05452088846100701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,3584,0.020407110452651978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,3072,0.02701955537001292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,3584,0.044403556320402354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,2560,0.023448000351587932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,6144,0.03170133299297757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,3072,0.039487110243903265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,5120,0.027462222509913977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,2560,0.033923556407292686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,2048,0.015791111522250705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,4096,0.04960088928540548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,1536,0.015199111567603217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,1536,0.013440888788965015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,1024,0.01163644426398807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,12288,0.11294133133358425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,2560,0.016574222180578444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,768,0.009669333696365356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,7168,0.06333511405520968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,768,0.014863999353514777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,512,0.007688889073001013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,768,0.011410666836632622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,1536,0.023317333724763658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,512,0.010786666638321347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,3072,0.01864177816443973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,2048,0.019536889261669584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,256,0.010488000180986194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,2048,0.028654221031400893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,128,0.010048000348938836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,64,0.004906666775544484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,256,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,512,0.012832889126406776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,1024,0.012044444680213928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,128,0.005626666463083691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,32,0.009806222385830348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,65536,0.39205511411031085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,128,0.010098666780524785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,51200,0.28613776630825466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,65536,0.5949760013156468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,32,0.0053022222386466126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,16384,0.09761510954962836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,1024,0.017847999930381775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,256,0.010767999622556897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,51200,0.4593831168280707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,16384,0.05785777833726671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,16384,0.14904710981580946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,10240,0.06488533152474298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,12288,0.04636355572276645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,12288,0.11355733871459961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,10240,0.040687110688951277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,10240,0.09502044651243423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,8192,0.03016622198952569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,7168,0.04379822148217095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,64,0.00982844415638182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,6144,0.03545599844720628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,51200,0.18029333485497367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,7168,0.06804533137215509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,12288,0.07080888748168945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,5120,0.030497776137457952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,6144,0.05838488870196872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,5120,0.020593777298927307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,8192,0.050717333952585854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,65536,0.21684355205959746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,4096,0.017697778013017442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,3584,0.020395555429988436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,7168,0.027240888939963445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,3584,0.016534222496880423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,3584,0.03661422265900506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,3072,0.018007111218240526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,4096,0.02275644408331977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,3072,0.01516711049609714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,5120,0.049508445792728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,3072,0.03238666719860501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,2048,0.013939556148317126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,2560,0.013964444398880005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,4096,0.0410444438457489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,2560,0.02786311176088121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,8192,0.0766320029894511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,2048,0.0127688886390792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,1024,0.009039999710188972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,1536,0.01184177812602785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,768,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,1024,0.010777778095669217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,1024,0.014929778046078153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,512,0.006267555471923616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,768,0.010433778166770935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,2560,0.01612622208065457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,256,0.005516444643338521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,512,0.009766221874290043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,512,0.01073333372672399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,128,0.004570666700601578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,256,0.009019555317031013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,2048,0.023748444186316595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,64,0.004358222087224324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,128,0.009378666679064432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,128,0.008599999878141616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,32,0.004512000001139111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,64,0.008391111261314815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,768,0.0124177775449223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,65536,0.30307645267910427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,51200,0.2302631007300483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,65536,0.20742400487263998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,256,0.009373333719041612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,16384,0.07466666566001044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,65536,0.5020782152811686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,51200,0.16529511080847845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,12288,0.05638666947682699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,16384,0.056977775361802846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,51200,0.41812088754442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,16384,0.1302808920542399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,10240,0.04602933261129591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,12288,0.04532533221774631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,8192,0.03914133376545376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,10240,0.040347556273142494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,1536,0.01110222190618515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,8192,0.029323554701275293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,8192,0.06685066885418363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,1536,0.019523556033770244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,7168,0.026138666603300307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,7168,0.05933155616124471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,6144,0.02304800020323859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,32,0.008336000144481659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,6144,0.05126488871044583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,12288,0.09866666793823242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,4096,0.019762666689025033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,5120,0.020443555381562974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,5120,0.0433991120921241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,3584,0.017478222648302715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,4096,0.03632622294955783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,6144,0.02369511127471924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,6144,0.02836444311671787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,3584,0.016569778323173523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,3584,0.03212533394495646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,2560,0.013401777380042605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,3072,0.01517599986659156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,5120,0.023192000057962205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,2048,0.01185955521133211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,2560,0.013968888256284924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,2560,0.024708444873491924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,1536,0.010030222435792288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,2048,0.012776888906955719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,2048,0.02109244465827942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,1024,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,1536,0.017522666189405654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,3072,0.01536355581548479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,768,0.006842666616042455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,1024,0.01073333372672399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,3072,0.028560890091790095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,768,0.011420444481902652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,10240,0.08284711175494723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,7168,0.033302221033308245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,4096,0.017679999272028606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,512,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,1536,0.011757333245542316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,128,0.004565333326657613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,256,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,256,0.00815555536084705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,1024,0.013449778159459433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,128,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,128,0.007661333514584436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,768,0.010386666489972008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,512,0.0058400001790788435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,64,0.0074622225430276655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,65536,0.20451822545793322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,32,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,256,0.004898666507667965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,51200,0.15600800514221191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,65536,0.4307564364539252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,51200,0.12941511472066244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,51200,0.33651823467678493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,32,0.004565333326657613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,12288,0.03880000114440918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,16384,0.04514133267932468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,16384,0.10978755686018203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,10240,0.03207022282812331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,12288,0.036610666248533465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,512,0.009702222214804756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,12288,0.0835262205865648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,8192,0.023074666659037273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,65536,0.16236800617641872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,7168,0.02032444377740224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,10240,0.07025155756208631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,6144,0.01726222203837501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,64,0.004221333397759331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,7168,0.020423110988405015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,5120,0.0151582227812873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,6144,0.018087110585636564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,6144,0.043503112263149686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,5120,0.016232000456915963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,5120,0.03707733419206407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,10240,0.0323884454038408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,4096,0.030928889910380047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,3584,0.01148888882663515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,8192,0.022630222969584998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,3072,0.010372444159454769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,16384,0.04920266734229194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,3584,0.027594667341974046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,7168,0.04982133375273811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,2560,0.009290667043791877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,3072,0.024504888388845656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,2048,0.008195555872387355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,2560,0.011930666863918304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,2560,0.02128355536195967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,1536,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,4096,0.014261333478821648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,1536,0.010395555860466426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,1536,0.01514844430817498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,8192,0.056624001926845975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,1024,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,768,0.005072888814740711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,4096,0.012495111260149213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,768,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,2048,0.01128266668981976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,2048,0.018294221825069852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,512,0.00872444444232517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,1024,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,3584,0.013452444639470843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,3072,0.012772444221708508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,128,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,256,0.007318221860461765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,128,0.008349333372380998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,64,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,128,0.006673777682913675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,32,0.003559999995761447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,64,0.006640888750553131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,512,0.008772444393899705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,32,0.006637333167923822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,256,0.008656000097592672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,1024,0.011740444435013665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,51200,0.09839289055930243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,65536,0.16000710593329534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,65536,0.33972711033291286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,768,0.010112889111042023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,16384,0.03837066557672288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,51200,0.12831821706559923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,51200,0.2692506578233507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,16384,0.09035999907387628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,256,0.003921777837806278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,12288,0.03654133280118307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,12288,0.06913333468967013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,8192,0.01683111157682207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,10240,0.05730933613247342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,8192,0.022170666191313002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,512,0.004576888763242298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,8192,0.046725332736968994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,12288,0.027093332674768236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,16384,0.04562755425771078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,7168,0.02036711076895396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,10240,0.02239377796649933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,5120,0.01109688894616233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,6144,0.0362737774848938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,10240,0.032582223415374756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,4096,0.009690666364298927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,7168,0.014527110589875115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,4096,0.025470222036043804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,4096,0.014434667097197639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,7168,0.04126222266091241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,65536,0.12236089176601833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,3072,0.008325333396593729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,5120,0.016930666234758165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,3072,0.020311110549502902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,3072,0.013125333521101209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,3584,0.009008888569143083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,6144,0.01851466629240248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,3584,0.013457777599493662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,2048,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,5120,0.03068355719248454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,2560,0.007687110867765214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,1536,0.005963555640644497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,6144,0.012757333616415659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,1536,0.010581333604123859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,1024,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,3584,0.02312711046801673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,2560,0.011756444142924415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,768,0.00444444434510337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,1024,0.010568888651000129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,2048,0.011151110960377587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,768,0.009370666411187913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,768,0.008998221821255153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,512,0.008671999805503422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,256,0.0038968887594011095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,512,0.007690666450394525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,1536,0.012808889150619507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,128,0.003208000005947219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,2560,0.01770400007565816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,256,0.006274666637182236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,128,0.008349333372380998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,128,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,64,0.0057653333577844836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,2048,0.015336889359686108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,512,0.004222222086456087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,65536,0.10029244422912598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,51200,0.0804435544543796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,65536,0.15997689300113255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,65536,0.301162666744656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,256,0.008341333104504479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,16384,0.030613332986831665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,51200,0.1321457756890191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,51200,0.2415928840637207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,1024,0.009754666851626502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,32,0.003256888853179084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,16384,0.04641333222389221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,10240,0.01588888963063558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,12288,0.037977778249316745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,8192,0.013748444616794586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,10240,0.03311022122701009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,10240,0.05063555637995402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,7168,0.012435555458068848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,8192,0.04104355639881558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,64,0.0031964443624019623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,6144,0.01106844428512785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,16384,0.0789804458618164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,12288,0.06110044320424398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,6144,0.01977422171168857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,5120,0.010244444840484196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,8192,0.02369155486424764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,5120,0.01756355497572157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,4096,0.008913777768611908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,5120,0.027475555737813313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,3584,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,4096,0.01516355574131012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,7168,0.021094222863515217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,4096,0.022844443718592327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,3584,0.013855111267831592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,3584,0.02055288851261139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,2560,0.006637333167923822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,3072,0.01826400061448415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,6144,0.03182844320933024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,2048,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,2560,0.0124195557501581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,2560,0.016366221838527255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,1536,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,2048,0.014146667387750415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,7168,0.03605155481232537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,3072,0.007127111156781514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,1536,0.01072266697883606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,1024,0.009696000152164036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,768,0.003956444561481476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,32,0.005836444596449534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,768,0.009358222285906473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,768,0.00867733359336853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,2048,0.011408000356621213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,512,0.008678221868144141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,256,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,1536,0.011922666596041786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,256,0.008384889198674096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,256,0.00591822216908137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,128,0.005299555758635203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,128,0.008368888662921058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,64,0.003198222153716617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,512,0.007296889192528195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,32,0.0032071111102898917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,3072,0.013136000268989138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,128,0.0033191111352708605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,65536,0.0952942238913642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,32,0.0053182223604785065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,512,0.0038968887594011095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,16384,0.028774221738179524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,65536,0.29988622665405273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,12288,0.018950222267044913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,51200,0.2347013420528836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,1024,0.004579555657174853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,16384,0.07868178023232354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,12288,0.060079110993279346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,8192,0.012768000364303589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,10240,0.05032889048258463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,7168,0.011400000088744693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,8192,0.04121333360671997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,6144,0.010307555397351583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,7168,0.03603733248180813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,5120,0.009008000294367472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,6144,0.03134311238924662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,4096,0.00795555528667238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,5120,0.026996443669001263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,51200,0.07003555695215861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,4096,0.022647110952271357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,3072,0.006609777609507243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,3584,0.02029600077205234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,2560,0.006255110932721033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,3072,0.01831288966867659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,2048,0.005232000102599462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,1024,0.009704888694816166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,2560,0.01591199967596266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,1024,0.004234666625658671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,2048,0.014072888427310519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,768,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,1536,0.011857777833938599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,512,0.003587555554178026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,768,0.00832444429397583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,256,0.0031928889867332247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,12288,0.020957332518365648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,128,0.0030453333424197305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,64,0.005250666704442766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,256,0.005906666732496685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,64,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,32,0.0029128889242808023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,3584,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,65536,0.09029599693086411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,1024,0.009425777528021071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,32,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,51200,0.0722000002861023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,16384,0.026300445199012756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,10240,0.014466666513019137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,65536,0.29940443568759495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,51200,0.2411573463016086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,1536,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,128,0.005326222214433882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,8192,0.011100444528791638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,64,0.005274666680230035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,10240,0.05021689004368252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,7168,0.009696000152164036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,6144,0.009352000223265754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,8192,0.040863109959496394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,5120,0.011762667033407422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,512,0.007358222372002072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,12288,0.017557332913080852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,6144,0.03144266539149814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,3584,0.008662222160233391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,16384,0.08033955759472318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,5120,0.027170666389995154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,3072,0.008667555948098501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,2560,0.007970666719807519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,3584,0.02086399992307027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,2048,0.005548444473081165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,3072,0.018741332822375827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,1536,0.00462400002612008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,7168,0.03588355580965678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,1024,0.003891555385457145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,1536,0.012126222252845764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,768,0.003858666867017746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,10240,0.012778666284349231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,512,0.003530666646030214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,1024,0.009476444787449306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,256,0.0031600000543726813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,768,0.008394666843944127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,128,0.0028880000528362063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,512,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,64,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,256,0.005853333406978183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,32,0.0029075555503368378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,128,0.0053084443012873335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,4096,0.010076444182131026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,32,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,12288,0.05951199928919474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,2560,0.016532444291644625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,2048,0.014141332772043018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,64,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,4096,0.02252799934811062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,16384,1.750666618347168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,16384,5.425482855902778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,16384,5.904585944281684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,51200,6.673920101589626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,12288,4.076898786756727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,16384,5.390047285291884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,12288,1.1605973773532443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,12288,4.126649644639757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,10240,3.3134312099880643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,10240,3.430426703559028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,10240,1.0624817742241752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,12288,3.7798737419976125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,8192,0.7937128808763293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,8192,2.629415088229709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,8192,2.722215016682943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,10240,3.0694569481743703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,8192,2.515334235297309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,51200,16.20556640625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,7168,2.2653564876980252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,7168,0.7035377820332845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,7168,2.343517303466797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,6144,0.6202995512220595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,6144,1.9533244238959417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,6144,2.109459559122721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,7168,2.180302302042643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,5120,0.533100446065267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,6144,1.9435910118950739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,4096,1.3665812810262044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,5120,1.6287511189778645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,5120,1.8309618631998699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,4096,0.46411911646525067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,4096,1.2583235634697807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,3584,1.0627750820583768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,4096,1.271996392144097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,3584,0.40372443199157715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,5120,1.5201217863294814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,3584,1.1015155580308702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,3072,0.8911138110690646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,3072,0.349451568391588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,3072,0.965452406141493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,2560,0.8178062438964844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,3072,0.9510889053344727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,2560,0.3047155539194743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,2560,0.8137795660230848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,3584,1.0798639721340604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,2048,0.7057653533087836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,2560,0.7467262479994031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,2048,0.2552106645372179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,51200,24.337014092339412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,1536,0.47498130798339844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,2048,0.6135208871629503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,1536,0.20554578304290771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,1024,0.3498373296525743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,1536,0.5570613543192545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,1536,0.48074134190877277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,1024,0.16262400150299072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,768,0.28369066450330943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,768,0.14194844828711614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,512,0.22386489974127877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,2048,0.6700346734788684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,512,0.12598489390479192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,512,0.2597457832760281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,256,0.14781333340538874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,1024,0.39929599232143825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,512,0.28030843204922146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,128,0.11372533109452988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,1024,0.368015103869968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,256,0.1869964467154609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,768,0.30232710308498806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,64,0.10034488969379002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,128,0.10932088560528225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,128,0.22090044286515978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,128,0.14266578356424967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,64,0.22774844699435762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,32,0.2323831187354194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,256,0.23425422774420845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,768,0.30611467361450195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,32,0.09895733329984878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,51200,18.12472873263889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,256,0.1106497777832879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,51200,5.002038319905599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,65536,7.20355224609375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,16384,4.583164638943143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,51200,12.292018466525606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,16384,5.109500461154513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,16384,1.2660737567477756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,65536,16.369394938151043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,12288,3.193471060858833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,16384,4.105092366536458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,12288,0.9758666356404623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,12288,3.366679085625542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,51200,14.167781405978731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,51200,19.22406684027778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,12288,2.9424976772732205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,10240,2.6564888424343533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,10240,0.7797982427808973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,10240,2.680110295613607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,10240,2.4825856950547958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,8192,0.6373510890536839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,8192,2.255239062839084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,8192,2.0874178144666886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,7168,0.6506960127088759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,7168,1.8667305840386286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,8192,1.960062238905165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,7168,1.8222177293565538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,6144,1.4883769353230794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,65536,24.601903279622395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,6144,0.4848480224609375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,6144,1.5162009133232965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,5120,0.4094542132483588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,5120,1.4442479875352647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,6144,1.4675199720594618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,4096,1.0011777877807617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,4096,0.34227466583251953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,5120,1.175349341498481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,4096,1.000509368048774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,65536,19.811499701605904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,4096,0.9565021726820203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,5120,1.2679519653320312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,3584,0.8210239940219455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,3584,0.31139911545647514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,3584,0.8715315394931369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,3072,0.2712639967600505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,3072,0.8394817776150174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,3584,0.8301226827833387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,3072,0.7420630984836154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,7168,1.7084301842583551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,2560,0.23431200451321074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,2560,0.6125146547953287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,2048,0.44106578826904297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,3072,0.7032142215304904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,2560,0.6365031136406792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,2048,0.21936978234185112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,2048,0.514093346065945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,2048,0.47871642642551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,1536,0.42020355330573184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,1024,0.25561687681410045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,1536,0.39083200030856663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,1024,0.13132088714175755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,768,0.22083377838134766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,2560,0.641432868109809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,1024,0.283040894402398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,512,0.15149866210089788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,768,0.11247200436062282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,1536,0.35052000151740176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,768,0.23348710272047254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,512,0.10077066553963555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,1536,0.16148711575402155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,512,0.18733778264787462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,128,0.08552889029184978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,256,0.18518933984968397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,256,0.14830311139424643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,1024,0.30851022402445477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,64,0.07633688714769152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,128,0.08795555432637532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,768,0.241338676876492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,32,0.07797511418660481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,128,0.17355822192298043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,256,0.10331110821829902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,64,0.179439107577006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,32,0.1855644385019938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,256,0.08854133552975124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,128,0.11333955658806695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,65536,2.1289342244466147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,512,0.208898663520813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,16384,1.2976818084716797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,51200,3.9438116285536022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,51200,1.5041973325941298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,65536,4.888903299967448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,16384,0.41680532031589085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,12288,0.9838586383395724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,16384,1.329393810696072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,51200,6.31932873196072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,12288,0.32519910070631236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,12288,1.0558346642388237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,16384,1.467940436469184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,51200,4.347947438557942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,12288,0.8914000193277994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,10240,0.27305600378248424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,65536,8.000638326009115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,65536,6.275254143608941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,8192,0.7229910956488715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,8192,0.21280711226993135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,10240,0.7058017518785266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,8192,0.6866844495137533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,7168,0.1887448893653022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,8192,0.5916168954637315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,10240,0.9832222196790906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,7168,0.710417800479465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,6144,0.16872799396514893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,6144,0.4912675751580133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,5120,0.39334845542907715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,7168,0.6821075545416938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,5120,0.15131378173828125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,5120,0.4103999932607015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,6144,0.4210506545172797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,4096,0.32118577427334255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,4096,0.11842489242553711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,3584,0.27733511394924587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,4096,0.29737867249382866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,3584,0.10735733641518487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,3584,0.286352899339464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,3584,0.26589777734544545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,3072,0.2420266734229194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,7168,0.5012870894538032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,2560,0.19821955098046196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,3072,0.09427644146813287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,3072,0.22921154234144422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,6144,0.444339566760593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,2560,0.08254488971498278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,2560,0.21101510524749756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,2560,0.19498133659362793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,5120,0.3633875581953261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,2048,0.07003288798862033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,1536,0.12775644991132948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,2048,0.16236621803707546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,3072,0.24876978662278917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,10240,0.8674906624688042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,1024,0.09999733501010472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,1536,0.05699466996722751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,1536,0.13868978288438585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,2048,0.17508355776468912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,1536,0.13112800651126438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,768,0.06870933373769124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,1024,0.048447112242380776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,1024,0.10378577974107529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,1024,0.09950933191511367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,768,0.04413422279887729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,768,0.08351911438835992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,768,0.08187377452850342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,4096,0.3306071228451199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,512,0.07093955410851373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,512,0.04007110993067423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,512,0.0692888895670573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,128,0.030851556195153132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,256,0.03619555632273356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,256,0.05344622333844503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,128,0.05991022454367744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,128,0.035084442959891424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,128,0.043059554364946157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,32,0.026159110996458266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,256,0.03650666607750787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,64,0.061517331335279674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,32,0.06358044677310519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,256,0.06370755698945788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,512,0.05827022261089749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,64,0.02555999987655216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,2048,0.17457689179314506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,65536,1.716311984592014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,51200,3.06748538547092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,16384,1.1221662097507052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,51200,1.175530645582411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,65536,4.065184699164496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,16384,0.3169875674777561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,12288,0.7960835562811958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,16384,1.132229381137424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,51200,3.2544061872694225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,16384,0.9519199795193143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,12288,0.2462115552690294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,51200,4.866866641574435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,65536,4.5079833136664496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,10240,0.639686213599311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,12288,0.6893102327982584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,10240,0.2094942198859321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,65536,6.450976053873698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,10240,0.554074658287896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,8192,0.16852000024583605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,12288,0.8132906489902072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,7168,0.4591369099087185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,8192,0.44823646545410156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,8192,0.7326595518324109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,7168,0.45150846905178493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,7168,0.14765332804785833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,10240,0.6623653305901421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,6144,0.434754663043552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,6144,0.3776871098412408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,6144,0.32761777771843803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,5120,0.33084623018900555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,5120,0.11321600278218587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,5120,0.3114559915330675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,5120,0.28009867668151855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,4096,0.26208533181084526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,4096,0.0917137795024448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,3584,0.20543466673956978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,4096,0.2313777870602078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,6144,0.1321786642074585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,7168,0.3950657844543457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,3584,0.08245599932140775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,3072,0.19104444980621338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,3584,0.21983199649386934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,3584,0.2004293335808648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,2560,0.1505404445860121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,3072,0.19089066982269287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,4096,0.2453697787390815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,2048,0.12242844369676377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,2560,0.15187733703189427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,2048,0.13466044267018637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,2048,0.1265217728084988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,1536,0.09504977862040202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,8192,0.5854124493069118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,3072,0.07299555672539605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,1536,0.04526311159133911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,1024,0.07108800278769599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,3072,0.18080443806118437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,2560,0.1626933283276028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,768,0.05686399671766493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,1024,0.03857422206136916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,1024,0.08055111434724596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,1024,0.07812266879611544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,2048,0.05459200011359321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,768,0.03567644291453891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,768,0.06477155288060506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,768,0.06495555241902669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,256,0.030026667647891577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,512,0.05540711018774244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,512,0.05359911256366306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,2560,0.06519466638565063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,256,0.0494044456217024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,256,0.02955733405219184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,128,0.024285333024130926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,256,0.042340444193945996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,512,0.03964889049530029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,128,0.029176000091764662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,128,0.04651644494798449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,32,0.021309332715140447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,512,0.03284622232119242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,64,0.04774577750100029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,1536,0.10365866952472264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,64,0.02100444502300686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,128,0.03484800126817491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,1536,0.10716799894968669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,32,0.04888711041874356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,65536,1.2486018074883354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,51200,1.0003102620442708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,65536,3.586886935763889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,51200,2.5729698605007596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,16384,1.1002595689561632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,16384,0.9566159778171115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,16384,0.2702568901909722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,12288,0.741305775112576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,16384,0.7913671069675021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,65536,3.5720240275065103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,12288,0.654512882232666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,12288,0.20657866530948213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,51200,2.8158658345540366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,10240,0.5939857694837782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,10240,0.17570043934716117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,8192,0.4416826566060384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,10240,0.5692382388644749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,10240,0.47002845340304905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,8192,0.1609075599246555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,65536,4.881816016303168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,7168,0.37073598967658145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,8192,0.610038227505154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,8192,0.37073956595526797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,51200,3.8606686062282987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,7168,0.12524799505869547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,6144,0.31941244337293834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,12288,0.5651582082112631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,6144,0.11263022157880996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,5120,0.24002223544650605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,6144,0.32164178954230416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,6144,0.2817831039428711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,5120,0.2655342155032688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,4096,0.21216000450981987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,4096,0.07786044147279528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,4096,0.19379377365112305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,7168,0.3693466716342502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,7168,0.3350835641225179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,3584,0.18445066610972086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,3584,0.17224533028072783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,5120,0.23696798748440215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,3072,0.1610835525724623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,3072,0.06259466542137994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,3072,0.15106933646731907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,3584,0.16854222615559897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,2560,0.05490933193100823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,2560,0.13811288939581975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,2048,0.10352533393436009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,3584,0.07185155815548368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,4096,0.17638577355278862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,2048,0.0477128889825609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,1536,0.07951200008392334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,2048,0.11467022365993923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,2048,0.10887910922368367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,1024,0.05591289202372233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,2560,0.11503556039598252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,1536,0.0911324421564738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,1536,0.08851289086871678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,768,0.04550133479966057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,1024,0.06880622439914279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,1024,0.06913777854707506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,3072,0.16140711307525635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,768,0.05563377671771579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,768,0.05689866675270928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,1536,0.03991644581158956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,512,0.04670133193333944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,256,0.024534222152498033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,2560,0.12968356079525417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,5120,0.09656266371409099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,256,0.04233955674701267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,768,0.0317573348681132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,128,0.02247911029391819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,256,0.026198221577538386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,256,0.0370782216389974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,64,0.019359111785888672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,512,0.04743466774622599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,128,0.03963377740648057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,32,0.020085333122147452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,128,0.030341333813137476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,512,0.028691556718614366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,1024,0.034120000070995755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,64,0.04087555408477783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,32,0.041329777903027005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,512,0.0365955564710829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,128,0.025640888346566096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,65536,1.156183984544542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,65536,2.621831046210395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,51200,0.9071786668565539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,16384,0.6685591273837619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,51200,2.2695164150661893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,65536,2.889926274617513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,51200,3.431120978461372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,51200,2.2227164374457464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,16384,0.8919600380791558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,16384,0.24859378072950575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,65536,4.558228386773004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,12288,0.18979733520083955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,12288,0.540770689646403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,10240,0.4141857888963487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,12288,0.5241093105740017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,10240,0.14904622236887613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,8192,0.3160586622026232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,10240,0.4369155565897624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,8192,0.1195217768351237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,7168,0.29165156682332355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,8192,0.4585564401414659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,8192,0.2968177795410156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,7168,0.10907999674479167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,16384,0.6303395695156521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,6144,0.2519413365258111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,7168,0.26803376939561635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,6144,0.09198666943444146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,5120,0.2030266655815972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,6144,0.25668711132473415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,12288,0.46651289198133683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,6144,0.22852532068888345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,5120,0.21767466598086885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,5120,0.19109511375427246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,4096,0.171707550684611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,3584,0.14662578370836046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,4096,0.1593502230114407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,3584,0.059691554970211454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,3072,0.11958400408426921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,4096,0.16603022151523167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,10240,0.377271122402615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,4096,0.06781688663694593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,3072,0.13247556156582302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,7168,0.301383097966512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,2560,0.1064293318324619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,2560,0.04734222094217936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,3584,0.15204711755116782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,2560,0.11394400066799587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,2048,0.08180177874035306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,3584,0.14021510548061794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,5120,0.08037955231136747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,2048,0.09454666905932957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,2048,0.09125600258509318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,3072,0.053222222460640796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,1536,0.0751039981842041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,3072,0.1270622279908922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,1536,0.07321421967612372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,1024,0.02886755598915948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,768,0.03345422281159295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,1024,0.05753155549367269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,1536,0.06073421902126736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,768,0.045765333705478244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,512,0.02500444485081567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,1024,0.049032890134387545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,512,0.03806666533152262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,1024,0.05671022335688273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,512,0.03916710946294997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,256,0.01944533321592543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,2048,0.03998400105370416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,768,0.025527111358112756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,256,0.02132622235351139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,128,0.016154666741689045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,1536,0.03394755721092224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,128,0.0214044451713562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,64,0.014300444059901767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,512,0.022766222556432087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,2560,0.1075893309381273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,32,0.015142222245534262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,64,0.031649778286616005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,256,0.03341689043574863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,32,0.031545778115590416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,768,0.04799555407630073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,128,0.03132266799608866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,128,0.02641600039270189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,65536,0.9644951290554471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,256,0.030605332718955144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,16384,0.6572675704956055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,51200,2.250266604953342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,51200,0.763997342851427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,65536,2.92193963792589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,16384,0.2181697818968031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,16384,0.6678533554077148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,65536,3.4256657494439016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,51200,2.017118242051866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,65536,2.628302256266276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,12288,0.15699377324846056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,12288,0.5040159755282932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,10240,0.3893057770199246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,12288,0.4212808873918321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,10240,0.44461599985758465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,8192,0.33328978220621747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,10240,0.342137787077162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,12288,0.5550142394171821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,51200,2.8226426442464194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,8192,0.10616533623801337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,8192,0.3781235482957628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,7168,0.26847465833028156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,7168,0.09440444575415717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,7168,0.2414177788628472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,6144,0.21155822277069092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,6144,0.0866675575574239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,16384,0.582272900475396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,8192,0.3085973262786865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,6144,0.2088622252146403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,5120,0.17731822861565483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,4096,0.12249244583977593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,5120,0.07185333304935031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,5120,0.1914364496866862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,10240,0.13343377908070883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,5120,0.17659289307064482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,3584,0.10536444187164307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,4096,0.15399021572536892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,4096,0.14544533358679876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,3584,0.05494844582345751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,3072,0.10044355524910821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,7168,0.24520089891221789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,3584,0.12984266546037462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,2560,0.0750088890393575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,3072,0.048150221506754555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,3072,0.11834932698143853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,3072,0.11401332749260797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,2560,0.04295733239915636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,2560,0.10168533192740546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,2560,0.098471999168396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,4096,0.060552888446384005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,2048,0.03640000025431315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,2048,0.0843226644727919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,2048,0.08330844508277045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,6144,0.23129155900743273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,1536,0.03030666708946228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,1536,0.06772622134950426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,1536,0.06754399670494927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,1024,0.025612443685531616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,1024,0.050473776128556996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,1024,0.051467554436789624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,2048,0.06082399686177572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,512,0.028600888119803533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,768,0.022950222094853718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,1536,0.060791108343336314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,768,0.043772445784674756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,256,0.02144177754720052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,512,0.03353510962592231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,512,0.020676443974177044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,1024,0.04427288969357809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,512,0.036222222778532244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,128,0.016928889685206942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,256,0.01889955500761668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,256,0.029262221521801416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,768,0.034099555677837796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,64,0.013953778478834363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,128,0.018956444329685636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,128,0.02780533168050978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,128,0.023835556374655828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,768,0.04169777697987027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,32,0.027814222706688776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,3584,0.1363848845163981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,256,0.028520888752407495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,32,0.01462399959564209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,65536,0.8675440152486166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,65536,1.955751207139757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,51200,1.6935226652357314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,16384,0.4927040206061469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,64,0.027775999572541978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,65536,2.95777956644694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,65536,2.280408011542426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,16384,0.18793066342671713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,16384,0.5144151051839193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,12288,0.3924426767561171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,51200,2.41960080464681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,51200,0.6794071197509766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,10240,0.33299732208251953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,12288,0.3604008886549208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,10240,0.34351287947760684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,8192,0.27384887801276314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,16384,0.5903111033969456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,10240,0.30111556582980686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,51200,1.7223084767659504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,7168,0.23933421240912545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,8192,0.2658070988125271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,12288,0.1378177801767985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,12288,0.4438017739189996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,6144,0.2145235538482666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,7168,0.23496799998813203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,7168,0.2128444380230374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,10240,0.12961955865224203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,6144,0.07331644164191352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,6144,0.1824071142408583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,8192,0.09337511327531602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,5120,0.06471288866466947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,8192,0.2479555606842041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,7168,0.08523822493023342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,4096,0.13483288553025988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,5120,0.16501066419813368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,4096,0.12740888860490587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,4096,0.053136888477537364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,5120,0.16767288578881157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,4096,0.1314364406797621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,5120,0.15439200401306152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,3072,0.10443644391165839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,3072,0.10434310966067845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,6144,0.19927910963694254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,3072,0.042621334393819175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,3584,0.11873777707417806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,3584,0.04815555612246195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,3584,0.11364177862803142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,2048,0.06821155548095703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,2560,0.08671910895241632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,3072,0.10089155700471665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,2048,0.0742808911535475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,2560,0.07777955796983507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,1536,0.05899733304977417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,2560,0.08972177902857463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,1536,0.06129333045747545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,1536,0.027447111076778833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,1024,0.04521422253714668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,2048,0.03176177872551812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,1536,0.05343111024962532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,768,0.02872266703181797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,768,0.036320000886917114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,1024,0.03437599870893691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,2560,0.03791288865937127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,768,0.03871555460823907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,512,0.02998488810327318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,512,0.032637334532207914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,256,0.01641244524055057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,1024,0.02308799988693661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,256,0.025916443930731878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,512,0.021910222040282354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,256,0.02513955533504486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,128,0.013677332964208392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,128,0.024671110841963027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,128,0.021300445000330608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,512,0.018430221411916945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,64,0.012485333614879184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,1024,0.04639288783073425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,32,0.01277955538696713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,64,0.024421332610978022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,32,0.024531554844644334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,768,0.020065777831607394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,256,0.017076444294717576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,3584,0.11921421686808269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,128,0.016883555385801528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,65536,0.7017697758144803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,65536,1.604341295030382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,51200,1.4151093165079753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,16384,0.42898045645819766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,51200,0.5739635361565484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,16384,0.1685333384407891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,65536,2.51201417711046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,16384,0.41905421680874294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,51200,1.4388799667358398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,12288,0.12963199615478516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,12288,0.3068266709645589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,10240,0.2825431029001872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,65536,1.9074249267578125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,12288,0.330951107872857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,51200,2.1295750935872397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,10240,0.1026373306910197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,10240,0.2539786762661404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,2048,0.07271999782986112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,8192,0.08214222060309516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,8192,0.22548445065816244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,7168,0.2045760022269355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,8192,0.20290754901038277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,7168,0.07374222411049737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,7168,0.19792355431450737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,7168,0.18086844020419648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,8192,0.2365199989742703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,6144,0.17311199506123862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,6144,0.06520266665352716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,12288,0.34993423355950254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,6144,0.15864355034298366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,4096,0.10645244518915813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,5120,0.14516711235046387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,5120,0.13304000430636936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,6144,0.17702755663130018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,4096,0.11547822422451443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,10240,0.292647123336792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,4096,0.11043200227949355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,3584,0.04059200154410468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,5120,0.13118310769399008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,3072,0.085043556160397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,3584,0.09864977995554607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,5120,0.05527911252445645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,2560,0.0653288894229465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,3072,0.036531554328070745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,16384,0.5354693200853136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,3072,0.08709510829713608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,3584,0.08942221932941014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,2048,0.05471733212471008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,2560,0.03244533472590976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,2560,0.07556177510155572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,3584,0.10236889123916626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,2048,0.0640524427096049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,1536,0.0426515539487203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,2048,0.06329955657323201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,1536,0.02292977770169576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,1536,0.050945778687795006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,1024,0.02991022335158454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,1536,0.05317866802215576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,4096,0.04615199897024366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,2560,0.0773600008752611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,1024,0.020243555307388306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,1024,0.040159109565946795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,2048,0.02749066717094845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,512,0.018086221483018663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,768,0.034550223085615374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,3072,0.08985333310233222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,512,0.025920889443821375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,256,0.013712888790501488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,512,0.01644711030854119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,768,0.022557333111763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,1024,0.03843466771973504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,128,0.012328888807031842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,256,0.015509333875444202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,256,0.021870222356584337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,768,0.0317191117339664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,64,0.010381333529949188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,128,0.02132533325089349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,128,0.01908266709910499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,64,0.021221333079867895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,32,0.020992888344658747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,256,0.022089777721299067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,768,0.01850488947497474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,32,0.011735999749766456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,512,0.028703110085593328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,51200,1.2075697580973308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,65536,2.0053369734022355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,65536,1.4975235197279189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,128,0.015206222732861837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,16384,0.36160000165303546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,51200,1.5403110716078017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,16384,0.13720799816979304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,16384,0.41987021764119464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,51200,1.1208960215250652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,16384,0.3394924534691705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,12288,0.11454044448004828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,12288,0.2859644360012478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,12288,0.25233954853481716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,10240,0.23550489213731554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,51200,0.4707671271430121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,10240,0.2061742279264662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,12288,0.27026134067111546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,8192,0.18715200159284803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,10240,0.23398311932881674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,7168,0.16170756022135416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,7168,0.167142219013638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,10240,0.087281776799096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,65536,0.5632053481207954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,7168,0.14753333727518717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,7168,0.06165155437257555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,6144,0.05510222249560886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,6144,0.14415021737416586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,8192,0.06965600119696723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,6144,0.1282328897052341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,8192,0.1678462160958184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,5120,0.047242667939927846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,5120,0.12147822644975449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,5120,0.11024977763493855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,65536,1.6572542190551758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,6144,0.14248889022403294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,8192,0.19277332888709175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,4096,0.0916631089316474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,3584,0.0859911110666063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,3584,0.08437333504358928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,4096,0.08968977795706855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,3072,0.07545777824189928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,3584,0.08000977834065755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,4096,0.039460443788104586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,3072,0.03105777833196852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,3584,0.03420622150103251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,2560,0.06474488973617554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,2560,0.06449600060780843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,2048,0.046718223227394946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,2048,0.054479112227757774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,2048,0.053435554107030235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,4096,0.09664889176686604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,3072,0.07378844420115153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,5120,0.10636888609992133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,3072,0.06540622313817342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,1024,0.02488800055450863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,1536,0.02058844433890449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,1024,0.03234044379658169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,1024,0.03394133183691237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,1536,0.03322044346067641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,1536,0.04308177696333992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,768,0.026591110560629103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,1536,0.04426666763093737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,2560,0.0281315545241038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,768,0.016068443655967712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,768,0.0287715560860104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,512,0.02216800053914388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,2560,0.05683288971583048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,256,0.012233778006500669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,512,0.014743111199802823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,512,0.014744000302420722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,256,0.018915555543369718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,256,0.013895110951529609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,128,0.010796444283591377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,2048,0.024104000793562994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,128,0.018264888061417472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,64,0.009696889254781935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,128,0.013779555757840475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,32,0.010246222217877706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,64,0.01815822223822276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,32,0.017878222796652053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,256,0.0199937770764033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,1024,0.017243555850452848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,128,0.017320000463061862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,65536,0.503694216410319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,65536,1.194467544555664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,51200,0.9293795691596137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,51200,0.3912151124742296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,768,0.019630221856964957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,65536,1.7175120247734919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,16384,0.31796354717678493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,65536,1.4186577267116969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,512,0.023912888434198167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,51200,1.3080399831136067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,16384,0.12468088997734918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,12288,0.24518132209777832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,16384,0.3605528937445746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,12288,0.09360177649392022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,16384,0.3272213406032986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,12288,0.271558231777615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,8192,0.16131999757554796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,10240,0.21130043930477568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,10240,0.2068542242050171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,8192,0.06344711118274264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,7168,0.13777421580420599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,8192,0.16388977898491755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,51200,1.086992899576823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,10240,0.20895022816128203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,7168,0.14574133025275335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,10240,0.08033866352505155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,6144,0.12837866942087808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,6144,0.05043022169007195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,8192,0.16994667053222656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,5120,0.09727199872334798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,5120,0.043086220820744835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,5120,0.1088213324546814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,4096,0.07478577560848661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,5120,0.10815733008914524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,6144,0.12296266025967068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,3584,0.06474310821957059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,12288,0.24301687876383463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,4096,0.09085244602627224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,6144,0.12690311008029515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,3584,0.03271199928389655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,3072,0.059655997488233775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,3584,0.07956177658504911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,3072,0.028563555743959215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,7168,0.14959643946753606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,2560,0.046494222349590726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,3072,0.07074044148127238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,4096,0.03512533174620734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,4096,0.08813333511352539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,2560,0.0581644442346361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,2560,0.0609475572903951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,3584,0.07881777816348605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,2048,0.048956443866093956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,1536,0.030610667334662542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,2048,0.051129778226216636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,1536,0.019488000207477145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,3072,0.06850577725304498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,1024,0.02050844497150845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,1536,0.04251199960708618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,2560,0.025855110751258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,2048,0.04038133223851522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,768,0.017245334055688646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,1024,0.03237333231502109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,768,0.015523556205961438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,768,0.024031110935741003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,768,0.027686221732033625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,7168,0.056919998592800565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,1536,0.039444443252351545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,256,0.010374222364690568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,512,0.014103111293580798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,1024,0.016543999314308167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,1024,0.029284444120195176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,128,0.009385777844323052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,256,0.013328889177905189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,256,0.01920977731545766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,2048,0.0221413340833452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,512,0.013841778039932251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,64,0.00871022211180793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,128,0.0131217779384719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,128,0.016699555847379897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,64,0.01620533400111728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,512,0.023242667317390442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,32,0.01629955569903056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,256,0.017289777596791584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,65536,1.1401146782769096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,128,0.016286222471131217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,32,0.010063110954231685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,65536,0.45125246047973633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,512,0.01997777819633484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,16384,0.2576844427320692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,65536,1.3677138222588434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,65536,1.2342187033759224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,51200,0.8866577678256564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,16384,0.11390933725568984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,16384,0.2975422276390923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,51200,0.9258587095472547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,12288,0.08950310945510864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,12288,0.23291020923190645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,12288,0.21515111128489176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,10240,0.07092889149983723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,10240,0.18900711006588408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,12288,0.22553777694702148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,51200,1.0427146487765842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,16384,0.2933502197265625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,8192,0.05750399827957153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,8192,0.15249421861436632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,8192,0.14676000012291804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,10240,0.1836177772945828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,7168,0.05150399936570061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,7168,0.13208533657921687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,6144,0.10907555288738674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,7168,0.12959467040167913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,5120,0.08191733227835761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,6144,0.04450133442878723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,6144,0.11452266905042861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,6144,0.11514488855997722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,4096,0.06194400125079685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,7168,0.12220444944169785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,5120,0.09637422031826443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,51200,0.35519732369316953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,4096,0.031756444109810725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,3584,0.061588446299235024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,10240,0.17917422453562418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,3584,0.06957866748174031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,3072,0.04925688770082262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,3584,0.028718223174413044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,5120,0.038089778688218855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,5120,0.09812266959084405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,2560,0.04235377907752991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,3072,0.026141333911154006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,3072,0.06302755408816867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,4096,0.07935377624299791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,4096,0.08028266827265422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,2560,0.051290667719311185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,2560,0.023815999428431194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,2560,0.055333332882987134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,8192,0.14245333936479357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,1536,0.026364444030655756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,2048,0.04405511087841458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,3584,0.07169510920842488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,2048,0.04668088754018148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,1024,0.019984000258975558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,1536,0.01811377704143524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,1536,0.03843911157713996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,3072,0.060093336635165744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,1024,0.026426666312747534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,1024,0.02946399980121189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,2048,0.03468622101677789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,512,0.01219288922018475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,768,0.02164711058139801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,768,0.02514044443766276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,2048,0.020417778028382193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,256,0.010712888505723743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,512,0.018066666192478605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,512,0.0211191111140781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,1536,0.03507555524508158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,128,0.009183111290136972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,256,0.01551644504070282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,1024,0.0155395550860299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,256,0.016535111599498324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,64,0.009928888744778102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,128,0.012814222110642327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,128,0.014839111102951897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,32,0.008623999853928884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,128,0.015965332587560017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,64,0.014506666196717156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,32,0.014669333895047506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,512,0.013456000222100152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,256,0.01258133351802826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,65536,0.7906595336066352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,65536,0.3838222291734483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,768,0.014575110541449653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,51200,0.714647134145101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,65536,1.1718693839179146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,16384,0.2461031013064914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,65536,1.0507058037651908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,16384,0.09805778000089858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,12288,0.17399822341071236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,51200,0.7885404692755805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,768,0.015375999940766228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,10240,0.14827199776967367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,12288,0.19386933909522164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,12288,0.19112888971964517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,51200,0.30784177780151367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,10240,0.16364888350168863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,51200,0.9230186674329969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,10240,0.15889067120022243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,16384,0.25485510296291775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,12288,0.07553778092066447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,8192,0.050311999188529126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,8192,0.13160978423224554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,7168,0.10212177700466579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,8192,0.1308871110280355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,6144,0.0818551116519504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,7168,0.11646489302317302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,7168,0.11403466595543756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,10240,0.06367466847101848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,16384,0.2799075444539388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,6144,0.03926488757133484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,5120,0.06631022029452853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,6144,0.09911021921369766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,4096,0.05696355634265476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,5120,0.08584000004662408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,5120,0.03369866808255514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,5120,0.08499377965927124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,3584,0.04953599969546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,4096,0.028280887338850234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,4096,0.0693733361032274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,3072,0.044756445619795054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,3584,0.06083466609319051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,3584,0.026012443833880957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,6144,0.09974666436513265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,3584,0.06311733192867704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,2560,0.033117334047953285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,3072,0.023397333092159692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,3072,0.053131557173199125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,2048,0.03028800090154012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,2560,0.020940444535679285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,2560,0.04621422290802002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,2560,0.04992266827159458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,1536,0.02161600026819441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,2048,0.03865155577659607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,2048,0.04159377680884467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,4096,0.07071733474731445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,1024,0.01621866722901662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,1536,0.031535999642478094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,8192,0.11594222651587592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,1536,0.03362133436732822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,768,0.012606222596433429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,1024,0.014067555467287699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,1024,0.0232933329211341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,3072,0.056362668673197426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,1024,0.02645066711637709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,768,0.013471110827393003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,2048,0.018497778309716117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,512,0.016307555966907077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,512,0.012480888929631976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,512,0.019362666540675692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,1536,0.016003555721706815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,256,0.011931555966536203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,128,0.007000000112586551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,256,0.016195555528004963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,512,0.010260444548394945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,768,0.01960088809331258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,64,0.006599111275540457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,7168,0.045588443676630654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,32,0.00666133314371109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,256,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,64,0.013065777719020844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,32,0.013100444442696042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,256,0.013945778210957846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,65536,0.6669511265224881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,128,0.011965333587593503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,768,0.02292266653643714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,128,0.015408888459205627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,65536,0.9324835671318902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,51200,0.5537688997056749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,65536,0.8296595679389106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,51200,0.2540622287326389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,16384,0.08306577470567492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,16384,0.24106844266255698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,51200,0.6707039939032661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,16384,0.21949511104159883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,128,0.013134222063753339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,10240,0.1154622236887614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,12288,0.16450666056738958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,12288,0.1608062187830607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,16384,0.17201244831085205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,10240,0.13953777154286703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,8192,0.0975688894589742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,51200,0.72198486328125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,12288,0.13337243927849665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,10240,0.13771466414133707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,65536,0.3268675539228651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,7168,0.0839208894305759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,8192,0.11094311210844253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,8192,0.04373777906099955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,12288,0.06439289119508532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,7168,0.03884711199336582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,7168,0.09883199797736274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,5120,0.05449600021044413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,6144,0.08595110972722371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,10240,0.055833776791890465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,5120,0.07261600097020467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,5120,0.07366400294833712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,8192,0.11307288540734185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,4096,0.043579555220074125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,6144,0.07003733184602526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,4096,0.02480799953142802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,4096,0.058223108450571694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,3584,0.04007555709944831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,4096,0.06126311090257433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,6144,0.0865795546107822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,3072,0.03447555502255758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,3584,0.02293511066171858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,3584,0.051651553975211255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,3584,0.055063999361462064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,2560,0.02942311101489597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,3072,0.04603999853134155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,3072,0.04898933238453335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,2560,0.0395084453953637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,7168,0.09874844551086426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,2048,0.03334844443533156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,3072,0.020741333564122517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,5120,0.029161777761247423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,2048,0.01665422154797448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,2048,0.036372443040212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,2560,0.019300444258583915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,1536,0.02683466672897339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,6144,0.0337119996547699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,1536,0.030436442957984075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,1024,0.01313333296113544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,1024,0.02036177780893114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,768,0.012102222277058495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,1024,0.023245332969559565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,768,0.012438221938080259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,768,0.016898666818936665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,512,0.010017777482668558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,768,0.020502222908867728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,1536,0.015191111299726697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,2048,0.02545333405335744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,256,0.007679999702506595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,512,0.017147555947303772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,256,0.011100444528791638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,128,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,256,0.013839999834696451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,1536,0.019300444258583915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,128,0.011448000040319232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,64,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,128,0.011055111057228513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,32,0.006764444626039929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,512,0.014156444205178155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,64,0.011409777734014722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,32,0.011161777708265515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,256,0.012096888489193387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,65536,0.5179964701334635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,65536,0.26922933260599774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,51200,0.4213271141052246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,512,0.011415111521879831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,65536,0.7662097613016764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,1024,0.01384622189733717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,65536,0.6868746545579699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,51200,0.20998933580186632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,51200,0.5963199933369955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,16384,0.0707262224621243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,12288,0.10775822401046753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,16384,0.17647643884023032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,128,0.013007111019558378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,12288,0.060421334372626416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,12288,0.13791999551984999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,10240,0.0952133337656657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,12288,0.13524177339341906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,16384,0.13716444704267713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,10240,0.05133777856826782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,10240,0.11706933710310195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,10240,0.11515643861558701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,16384,0.18398400147755942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,51200,0.5513928731282552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,8192,0.09317066934373643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,7168,0.03404088815053304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,7168,0.08282399839825101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,8192,0.07548977931340535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,6144,0.07150844732920329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,7168,0.06451200114356147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,5120,0.0450764430893792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,6144,0.03016266557905409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,6144,0.07192799780103895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,5120,0.026163554853863184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,5120,0.0607137746281094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,4096,0.03414933217896356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,5120,0.06207466787762112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,4096,0.04975377851062351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,4096,0.05176266696718004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,8192,0.039452443520228066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,3584,0.04429955614937676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,3584,0.04683199856016371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,3072,0.026537777649031744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,6144,0.0566142201423645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,3072,0.03921511107020908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,3584,0.031480888525644936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,8192,0.09512711233562893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,3072,0.041610668102900185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,3072,0.01794488893614875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,2560,0.03379822108480666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,2560,0.03613422314325968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,3584,0.019840889506869845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,2048,0.019096000327004325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,2048,0.028334223561816748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,2048,0.030464887619018555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,2560,0.024018666810459558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,2048,0.015344889627562629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,1536,0.02340000040001339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,1536,0.025482666161325242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,7168,0.08258222209082709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,1024,0.011760888828171624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,1024,0.011782222323947482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,1024,0.017663111289342243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,4096,0.02209866709179348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,1024,0.019882667395803664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,1536,0.014522666732470194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,512,0.008211555580298105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,768,0.015007111761305066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,768,0.01790488925245073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,512,0.012540444731712341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,256,0.006609777609507243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,2560,0.01663377715481652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,256,0.010072000324726105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,768,0.00944977750380834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,256,0.010495999621020423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,256,0.013473778135246701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,768,0.01124533348613315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,128,0.010072888599501716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,128,0.012832000023788877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,32,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,512,0.010601777997281818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,1536,0.013460444079505073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,32,0.009689778089523315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,65536,0.37926578521728516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,128,0.010027555955780877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,65536,0.22639555401272246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,65536,0.5579679807027181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,51200,0.2849217785729302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,512,0.01481244464715322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,16384,0.09715822007921006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,51200,0.45355643166436094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,128,0.006259555617968242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,51200,0.415845341152615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,12288,0.07424177726109822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,16384,0.14391999774509007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,64,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,16384,0.13902666833665636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,10240,0.06318400303522746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,12288,0.04571377899911669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,64,0.009776888622177972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,51200,0.16852888796064588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,10240,0.04053333401679993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,10240,0.09064888954162598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,16384,0.0572648909356859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,8192,0.0739591121673584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,8192,0.029669334491093952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,8192,0.07485244671503703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,65536,0.5307555728488499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,12288,0.10649510886934067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,7168,0.06490311357710096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,6144,0.03727466530270047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,10240,0.09250666697820027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,8192,0.046719110674328275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,5120,0.030086222622129653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,6144,0.0233262214395735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,6144,0.056332442495557994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,6144,0.0582266648610433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,5120,0.02052622205681271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,5120,0.04708800050947401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,5120,0.050029334094789296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,12288,0.10834755500157674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,3584,0.022376888328128394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,4096,0.039282666312323675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,7168,0.026359111070632935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,4096,0.04251377781232198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,3072,0.019707555572191875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,3584,0.01625422139962514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,3584,0.035396443472968206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,3584,0.038013332419925265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,2560,0.017070222232076857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,3072,0.031161778502994116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,4096,0.02436888880199856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,3072,0.033342222372690834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,7168,0.04101066788037618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,2048,0.01424888935354021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,2560,0.01386755539311303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,2560,0.029539555311203003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,2048,0.0229404436217414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,7168,0.06640088558197021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,2048,0.02511377798186408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,1536,0.011702222128709158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,1024,0.009352888498041365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,3072,0.015166223049163818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,1536,0.02130399975511763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,1024,0.010657777388890585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,768,0.008184000021881526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,2560,0.02679022153218587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,4096,0.017645332548353408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,1536,0.012095111111799875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,768,0.012307555311255984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,512,0.006656000183688269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,768,0.015214223000738354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,512,0.010517333116796283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,512,0.013162666724787818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,256,0.005586666779385672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,1024,0.01678311162524753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,2048,0.012606222596433429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,256,0.008792888787057664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,768,0.010207999911573198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,128,0.004565333326657613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,1536,0.018862222631772358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,128,0.009272889130645329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,64,0.004547555413511065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,1024,0.014477334088749357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,512,0.009729777773221334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,64,0.008047999607192146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,32,0.008044444852405125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,256,0.009371555513805812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,256,0.011071999867757162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,51200,0.22868445184495714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,128,0.008360889222886827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,65536,0.5018569098578559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,128,0.011042666931947073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,16384,0.07785244120491876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,32,0.004583111239804162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,51200,0.16416800022125244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,51200,0.3931902249654134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,16384,0.056688891516791455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,16384,0.13167999850379095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,65536,0.292193783654107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,16384,0.13815822866227892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,10240,0.0476835568745931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,12288,0.045124444696638316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,12288,0.0986293355623881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,65536,0.20841333601209852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,12288,0.10631733470492893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,8192,0.03854044609599643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,10240,0.03980355461438497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,10240,0.08381155464384292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,7168,0.03272622161441379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,65536,0.5263502332899306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,8192,0.02937155630853441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,8192,0.06744355625576444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,6144,0.027306665976842243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,12288,0.05910755528344048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,51200,0.4114346769120958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,7168,0.05915911330117119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,5120,0.02401155564520094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,6144,0.023051554958025616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,6144,0.051226665576299034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,6144,0.058098667197757296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,4096,0.01997955474588606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,5120,0.043297777573267617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,5120,0.0495013329717848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,10240,0.0897759993871053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,3584,0.01722400055991279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,4096,0.017510222064124215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,8192,0.07378577523761325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,4096,0.0414319998688168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,7168,0.026312889324294195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,3072,0.01603822244538201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,3584,0.01624000072479248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,3584,0.032241778241263494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,2560,0.01351199961370892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,3584,0.037243554989496865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,3072,0.015159999330838522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,3072,0.028692444165547688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,2048,0.011771555576059552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,5120,0.020115555988417733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,2560,0.024907555845048692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,2560,0.028720001379648846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,1536,0.0099973330895106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,4096,0.03628799981541104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,2048,0.02454400062561035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,1536,0.011747555600272285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,7168,0.06592533323499891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,1536,0.020865778128306072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,3072,0.032461331950293645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,2560,0.013884444203641681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,768,0.0070497774415545994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,1024,0.010739555789364709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,1024,0.016522667474216886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,2048,0.02126577827665541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,512,0.005741333381997214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,768,0.010213333699438307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,768,0.011416888899273343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,768,0.014832889040311178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,256,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,512,0.009743111001120673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,512,0.01276177747382058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,256,0.008266666697131263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,1024,0.01351111133893331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,256,0.010870222416188983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,2048,0.012726222475369772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,128,0.007673777639865875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,128,0.009264888862768808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,1536,0.017489777670966256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,1024,0.008016889293988546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,64,0.007459555235173967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,128,0.004254222330119875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,65536,0.20533333884345162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,64,0.003928889003064898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,512,0.009716444545321995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,65536,0.4312106768290202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,65536,0.16183377636803523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,256,0.009352000223265754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,51200,0.16148089038001165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,65536,0.3772275447845459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,16384,0.05193600058555603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,32,0.007318221860461765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,51200,0.33092623286777073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,128,0.010784000158309937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,16384,0.04491022229194641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,16384,0.10088621907764012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,12288,0.0362044440375434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,12288,0.0833502213160197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,10240,0.033634665939542986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,32,0.004038222134113312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,10240,0.032335999939176775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,10240,0.0703386664390564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,51200,0.12892799907260472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,10240,0.06669511397679646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,51200,0.2972559928894043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,8192,0.02203911046187083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,8192,0.05659377574920654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,8192,0.05518844392564562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,7168,0.020004444652133517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,12288,0.07807111077838473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,6144,0.019367110398080613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,7168,0.04913066824277242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,6144,0.018014222383499146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,5120,0.01717066599263085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,6144,0.04333066609170702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,12288,0.0421048899491628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,7168,0.02260533306333754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,5120,0.03721866673893399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,4096,0.013833777772055732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,5120,0.015862221519152325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,5120,0.03733333283000522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,3584,0.012704888979593912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,7168,0.05037066671583387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,4096,0.030981332063674927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,4096,0.031137777699364558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,3072,0.011735999749766456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,3584,0.013160000244776407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,6144,0.04342488778962029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,3072,0.012449777788586087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,2560,0.010392000277837118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,4096,0.01388533330625958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,2560,0.021313778228229944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,2560,0.011757333245542316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,8192,0.026610665851169165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,2048,0.00888000014755461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,3584,0.028018666638268366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,2048,0.011060444845093621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,2048,0.018224888377719455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,3072,0.025415110919210646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,2048,0.019128888845443726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,1536,0.010394666757848527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,1536,0.014959999256663851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,3584,0.027797334724002417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,2560,0.02240177823437585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,1024,0.011753777662913004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,16384,0.11001600159539117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,1024,0.009363555245929295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,1536,0.007634667058785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,1024,0.013350221845838757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,512,0.004588444613748127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,768,0.009224889179070791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,768,0.010102222363154093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,768,0.01221422188811832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,256,0.00423555572827657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,512,0.00889155517021815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,512,0.00866844422287411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,512,0.011049778097205691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,128,0.003903111235962974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,256,0.008380444513426887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,256,0.007325333025720384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,256,0.009357333183288574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,3072,0.024745777249336243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,64,0.0035342222286595237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,128,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,128,0.00906755526860555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,64,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,65536,0.12289778391520183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,32,0.006635555376609166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,1024,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,51200,0.10088533163070679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,65536,0.34023377630445695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,768,0.005585777676767773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,128,0.008359111017651027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,32,0.003548444559176763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,16384,0.03987377882003784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,51200,0.127146667904324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,51200,0.28358666102091473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,16384,0.09103111426035564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,12288,0.029876444074842665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,16384,0.09649778074688381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,65536,0.15894667307535806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,12288,0.03589866558710734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,12288,0.07465155257119073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,51200,0.2687777678171794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,10240,0.05710311068428887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,1536,0.016536888149049547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,10240,0.03080000148879157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,10240,0.06355644596947564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,12288,0.06804622544182672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,8192,0.04598666561974419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,10240,0.02204977803760105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,8192,0.05260355605019463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,8192,0.02182666626241472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,6144,0.014132445057233175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,7168,0.019271110494931538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,7168,0.04087911049524943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,7168,0.046682665745417275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,6144,0.03594311078389486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,6144,0.017616889543003507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,65536,0.3612231148613824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,7168,0.015998222761683993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,5120,0.03542666633923849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,5120,0.015641777051819693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,4096,0.025236444340811834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,16384,0.04418666826354133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,4096,0.029528889391157363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,3584,0.009952888720565373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,6144,0.04080266753832499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,3584,0.013120888835854001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,3584,0.023544000254737005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,3072,0.008769777913888296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,3584,0.02687377731005351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,4096,0.010374222364690568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,2560,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,3072,0.012442666623327466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,3072,0.02036977807680766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,3072,0.023743111226293776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,2048,0.0069511110583941145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,2560,0.01787644459141625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,5120,0.012465777496496836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,4096,0.013834666874673633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,1536,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,2048,0.011066666907734342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,2048,0.015452444553375244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,2048,0.018352889352374606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,1024,0.005381333331267039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,1536,0.01015466699997584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,1536,0.015578667322794596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,1024,0.010301333334710863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,768,0.004567111117972268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,1024,0.012813333008024426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,2560,0.011787555283970304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,8192,0.017858665850427415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,2560,0.021523555119832356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,768,0.009148444566461775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,768,0.008986666798591614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,768,0.011751111182901593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,5120,0.03057510985268487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,1536,0.012792888614866467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,512,0.010625777973069085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,256,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,1024,0.009467555416954888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,256,0.009001777403884465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,512,0.0042239998777707415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,128,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,64,0.0031884445084465873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,128,0.008375110725561777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,32,0.0032017777363459268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,256,0.003543111185232798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,65536,0.09349066681332058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,32,0.005758222192525864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,256,0.008362666600280339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,128,0.003532444437344869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,51200,0.0750088890393575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,65536,0.15778311093648276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,128,0.009027555584907532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,512,0.008678221868144141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,16384,0.030772444274690416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,64,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,51200,0.2801004515753852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,16384,0.045159111420313515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,12288,0.02239377796649933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,65536,0.3005777729882134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,65536,0.35610400305853945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,512,0.007630222373538547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,10240,0.01648977730009291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,12288,0.07334399885601468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,51200,0.1271386676364475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,10240,0.03188000122706095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,10240,0.05145511031150818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,16384,0.07869599925147162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,10240,0.06270933151245117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,16384,0.09529511133829753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,7168,0.012524444195959302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,8192,0.023015111684799194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,8192,0.0413973331451416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,8192,0.05194311009513008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,51200,0.23487554656134713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,6144,0.011431111229790581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,7168,0.03665333324008518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,12288,0.0363395545217726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,7168,0.045864890019098915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,5120,0.010016000105275048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,6144,0.018630221486091614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,6144,0.03155555658870273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,8192,0.013954665925767688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,4096,0.00869333330127928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,5120,0.027310222387313843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,5120,0.034336888127856784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,4096,0.02238133384121789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,12288,0.059953775670793324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,3584,0.008211555580298105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,3584,0.020516445239384968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,3584,0.026307556364271376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,6144,0.04042933384577433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,3584,0.013831111292044321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,5120,0.016413332687483895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,3072,0.01276622215906779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,3072,0.018442667192882962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,4096,0.014518222875065274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,7168,0.020264888803164165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,2048,0.00563822231358952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,2560,0.016185777054892648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,2560,0.021019554800457422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,2048,0.011185777684052786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,3072,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,1536,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,2048,0.018048889107174344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,1536,0.010392888552612728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,1536,0.011759999725553723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,3072,0.0236799998415841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,4096,0.02861600120862325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,2560,0.012099555797047086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,768,0.003882666842805015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,1024,0.009700444837411245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,1024,0.012786666552225748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,2048,0.014124444789356656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,768,0.008374222450786168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,768,0.011400889191362592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,2560,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,1536,0.0155102229780621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,512,0.010080888867378235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,512,0.008671999805503422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,1024,0.009568000005351173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,256,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,256,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,128,0.0028844444702068963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,512,0.0035253332720862497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,128,0.008341333104504479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,64,0.002862222285734283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,1024,0.004231111043029361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,32,0.0028817777832349143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,256,0.003555555724435382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,65536,0.08733599715762669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,32,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,51200,0.07138666841718885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,256,0.00867022242810991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,768,0.009218666288587783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,65536,0.2997048960791694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,128,0.005346666607591841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,512,0.007377777662542131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,10240,0.015594666202863058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,12288,0.06087911128997803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,64,0.005286222116814719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,8192,0.012477333347002665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,7168,0.011519110865063138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,10240,0.050366222858428955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,6144,0.010401777923107147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,8192,0.04137511054674784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,16384,0.028838220569822524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,5120,0.008984888593355814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,7168,0.036016888088650174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,6144,0.03160266743765937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,3584,0.007320889168315464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,51200,0.23453333642747667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,16384,0.07862844732072619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,4096,0.022406222091780767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,2560,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,3584,0.02050044470363193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,2048,0.005575110928879843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,3072,0.01830222209294637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,1536,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,2048,0.014150222142537435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,1024,0.004270222038030624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,12288,0.019275556008021038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,768,0.003910222401221593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,5120,0.026886221435334947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,3072,0.006612444503439798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,1024,0.009651555783218807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,256,0.003272888975010978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,768,0.008350222474998897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,128,0.0032257777121331957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,512,0.007320889168315464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,64,0.0028764444092909494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,4096,0.007679111427730984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,32,0.002867555452717675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,64,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,512,0.003546666767862108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,65536,0.0879982246292962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,32,0.0052382221652401825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,51200,0.07131733496983846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,2560,0.016209777858522203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,16384,0.02641155487961239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,51200,0.2411937713623047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,128,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,12288,0.016699555847379897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,10240,0.011740444435013665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,12288,0.059911111990610756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,8192,0.010388444695207808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,10240,0.05132711264822218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,7168,0.009735999835862054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,256,0.005710222240951326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,6144,0.009352000223265754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,8192,0.0411208901140425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,5120,0.01109955542617374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,7168,0.03592711024814182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,1536,0.011807111402352652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,6144,0.03164266546567281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,16384,0.08047644297281902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,3584,0.00904444439543618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,5120,0.02695288923051622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,4096,0.022679999470710754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,2560,0.007630222373538547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,3584,0.021135111649831135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,3072,0.018590221802393597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,1536,0.004575110971927643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,65536,0.2992497815026177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,2048,0.01422133379512363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,1536,0.011937778029176923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,768,0.0038986665507157645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,1024,0.009702222214804756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,3072,0.008038221961922115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,512,0.003383111208677292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,256,0.003216888962520493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,768,0.008345777789751688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,128,0.0028684443483750024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,512,0.007355555891990662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,64,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,2560,0.016304888659053378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,32,0.0028862222615215513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,128,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,4096,0.009677333136399588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,32,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,2048,0.005296000176005893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,256,0.005642666584915585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,1024,0.003922666526503033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,64,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,16384,1.6317413118150499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,12288,3.6595717536078562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,16384,4.919379764133029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,12288,1.203552034166124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,16384,5.956655290391709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,12288,4.108074612087674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,51200,6.514122856987847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,10240,1.0018302069769967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,10240,3.520478142632378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,8192,0.8179270956251355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,8192,2.8045342763264975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,8192,2.53726938035753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,7168,2.259126239352756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,10240,3.345424864027235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,7168,2.1709342532687717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,6144,0.6062649091084799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,7168,0.6857262187533908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,6144,2.1061920589870877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,5120,0.5216719839307997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,5120,1.5835030873616536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,6144,1.7907386355929906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,4096,1.251908408270942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,5120,1.540809737311469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,3584,1.0352986653645833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,4096,1.2134870953030057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,51200,15.701173570421007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,3584,1.088149282667372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,4096,0.4289404551188151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,3072,0.3367973433600531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,3072,0.893337779574924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,3584,0.38309423128763836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,2560,0.8545635541280111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,2560,0.2965048948923747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,2560,0.7646329138014051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,128,0.008346666892369589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,2048,0.6122248967488607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,2048,0.27155645688374835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,3072,0.9847840203179253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,1536,0.5034977595011393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,1024,0.35182666778564453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,1536,0.4925075636969672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,1024,0.37235376569959855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,768,0.13869067033131918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,512,0.22054666943020293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,512,0.15377867221832275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,2048,0.620398203531901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,512,0.24961778852674696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,256,0.18175644344753691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,1024,0.1616951094733344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,256,0.10918488767411973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,256,0.22473245196872285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,128,0.10820621914333767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,768,0.37676355573866105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,64,0.09437066978878444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,128,0.21184089448716906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,32,0.10015733374489678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,1536,0.19945866531795928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,64,0.21691377957661948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,768,0.2692364321814643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,32,0.21938667032453749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,128,0.12584711445702448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,51200,23.837541368272568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,16384,3.8065448337131076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,51200,4.864586724175347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,65536,6.614892323811849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,16384,4.46066411336263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,16384,1.2286222245958116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,12288,0.9630746841430664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,12288,3.411868413289388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,12288,3.1922124226888022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,10240,2.807956483629015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,10240,0.7762666808234321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,51200,12.162206861707899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,10240,2.517153845893012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,8192,2.0295768313937717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,8192,1.9551804860432942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,7168,0.5405306816101074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,7168,1.6469297409057617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,7168,1.9178123474121094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,6144,0.5051546626620823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,65536,15.869230482313368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,8192,0.708189328511556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,6144,1.4567279815673828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,5120,0.41226667828030056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,6144,1.4985876083374023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,4096,0.9855528937445747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,5120,1.2560524410671658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,4096,0.9247538248697916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,5120,1.2994950612386067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,3584,0.3119911087883843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,3584,0.7815422481960721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,3072,0.6920622189839681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,3072,0.2733688884311252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,51200,18.13739013671875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,2560,0.6001084645589193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,2560,0.23849066098531088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,4096,0.33923644489712185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,2048,0.4943982230292426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,2048,0.1993457741207547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,1536,0.3781840006510417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,2048,0.506147543589274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,1536,0.1623120043012831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,1024,0.25921334160698783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,1536,0.415521780649821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,768,0.2039946715037028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,1024,0.28947467274136013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,768,0.12946044074164495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,512,0.14990133709377712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,2560,0.612011538611518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,512,0.09991288847393459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,512,0.21213066577911377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,3072,0.7063111199273003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,128,0.0863386657502916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,768,0.22886755731370714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,256,0.10971466700236003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,128,0.16719288296169707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,64,0.07396088706122504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,32,0.07840444644292195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,256,0.08822044399049546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,64,0.17118667231665718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,32,0.17245154910617402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,128,0.08665244446860419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,3584,0.8677831225925021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,1024,0.12890311082204184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,256,0.1773395538330078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,65536,2.01947996351454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,51200,1.6462346182929144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,16384,1.1470008426242404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,65536,23.8783942328559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,16384,0.5503848923577203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,16384,1.4386782116360133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,12288,0.31270133124457467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,12288,0.8985075420803494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,65536,4.870368109809028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,12288,0.9918577406141492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,10240,0.2589084572262234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,10240,0.9252524905734592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,8192,0.2107715606689453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,10240,0.8199129104614258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,51200,3.834893544514974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,8192,0.6386666827731663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,7168,0.6817573441399468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,7168,0.19468177689446342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,6144,0.4747280014885797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,6144,0.19438221719529894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,5120,0.4084497822655572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,5120,0.1380364497502645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,4096,0.3046328756544325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,5120,0.3946097691853841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,51200,7.23468017578125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,4096,0.3129013379414876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,65536,8.216141594780817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,7168,0.5626399781968858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,8192,0.7507013214959039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,3584,0.11019021934933132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,6144,0.47527286741468644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,3072,0.23799554506937662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,2560,0.17284799946679008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,3072,0.23777511384752062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,2560,0.20057688819037545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,4096,0.11916266547309028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,2048,0.1450293329026964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,3584,0.2476960023244222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,1536,0.11641155348883735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,2048,0.07042222552829318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,2048,0.16554133097330728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,1536,0.05575288666619194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,3584,0.302656888961792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,768,0.07345244619581434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,1024,0.046712001164754234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,1024,0.09915022055308025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,512,0.046318223079045616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,768,0.0794008904033237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,512,0.039332442813449435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,256,0.037280887365341187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,1024,0.09063733286327785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,128,0.03162844313515557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,256,0.03553244471549988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,3072,0.09180800120035808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,256,0.0613973339398702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,64,0.028584887584050495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,32,0.03016888764169481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,2560,0.08012800084220038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,64,0.05907111035452949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,32,0.05940710835986667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,1536,0.13140355216132268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,128,0.03496355480617947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,768,0.04313955704371134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,512,0.06830400228500366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,65536,1.5530808766682942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,128,0.05769866704940796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,16384,0.8461795383029513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,51200,1.1240133709377713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,16384,0.32403466436598033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,12288,0.6944764455159506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,12288,0.2448240121205648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,65536,3.659204271104601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,16384,1.0760231018066406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,51200,2.921464072333442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,12288,0.7742844687567817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,10240,0.20400088363223604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,8192,0.16222933928171793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,10240,0.6074275440639919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,8192,0.49124887254503036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,7168,0.1498782237370809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,7168,0.38646843698289657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,7168,0.42232267061869305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,6144,0.3440213203430176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,6144,0.12473066647847493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,10240,0.5731973118252224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,5120,0.2831599977281359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,6144,0.3569475544823541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,51200,4.637752109103732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,4096,0.21554222371843126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,4096,0.09280800157123142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,3584,0.22173778216044107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,4096,0.25607821676466197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,3584,0.08320622311698066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,3584,0.20708089404635957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,3072,0.07342844539218478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,65536,6.928114573160808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,3072,0.18478666411505806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,2560,0.07226666476991442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,5120,0.10722844468222724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,2560,0.15431555112202963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,5120,0.3024142318301731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,1536,0.09320444530910915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,2048,0.1277466615041097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,3072,0.18315466245015463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,1536,0.10165511237250434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,2560,0.14382133218977186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,1024,0.03795199924045139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,2048,0.11656799581315781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,8192,0.505409770541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,512,0.03659466571278042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,768,0.03510488735304938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,2048,0.05343822307056851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,256,0.02825511164135403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,512,0.03227555420663621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,1536,0.04490844408671061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,512,0.05306933323542277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,128,0.024241778585645888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,256,0.047687109973695546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,128,0.028986665937635634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,768,0.050640000237358936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,128,0.044848889112472534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,32,0.020990222692489624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,64,0.045872890286975436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,32,0.046385778321160205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,1024,0.06447199980417888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,256,0.02903199858135647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,64,0.020610666937298246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,1024,0.0776488913430108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,768,0.06190222501754761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,65536,1.3055573569403756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,16384,0.7525964313083224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,51200,0.9694177839491102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,16384,0.26758400599161786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,65536,3.0687760247124563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,51200,2.5694683922661676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,16384,0.8670399983723959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,12288,0.20145422882503936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,10240,0.5197226736280653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,10240,0.17160356044769287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,10240,0.5209351115756565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,8192,0.42192800839742023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,8192,0.13866399394141302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,7168,0.36002667744954425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,8192,0.4416026539272732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,7168,0.12388177712758382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,51200,3.642856174045139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,6144,0.31037600835164386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,6144,0.10822933250003391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,5120,0.2495200104183621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,12288,0.6217599974738227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,6144,0.3050791157616509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,5120,0.09255022472805446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,4096,0.21263821919759116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,3584,0.16832978195614287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,4096,0.19994755585988364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,12288,0.6664613087972006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,3584,0.06902755631340875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,3072,0.1462577713860406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,2560,0.1278622282875909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,7168,0.38769690195719403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,3072,0.15352888902028403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,2560,0.05413066678576999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,2048,0.09745155440436469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,2560,0.13145689169565836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,4096,0.07872000005510119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,2048,0.10906933413611518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,3584,0.17750399642520479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,1536,0.03909600112173293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,1536,0.08690666490130955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,3072,0.06128888660007053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,768,0.04253333475854662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,1024,0.06567111280229357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,65536,5.16307110256619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,512,0.032925334241655134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,768,0.030866665972603693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,5120,0.2971511152055529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,256,0.024344000551435683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,512,0.028352889749738906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,512,0.0450799994998508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,256,0.02545333405335744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,256,0.040582223070992365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,1024,0.03326844506793552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,128,0.025302223033375208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,64,0.018925334016482036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,32,0.019945777124828763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,768,0.05343733231226603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,1536,0.07714311281840007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,32,0.0390035543176863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,1024,0.05296711126963297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,128,0.03790844480196635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,2048,0.04702577657169766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,65536,1.1573706732855904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,128,0.021498666869269475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,51200,0.8515991104973687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,65536,2.5947110917833114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,64,0.038889778984917536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,16384,0.695714685651991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,12288,0.18016000588734946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,16384,0.2461635536617703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,12288,0.4747609032524957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,12288,0.5014826456705729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,65536,3.7170863681369357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,10240,0.1455004480150011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,8192,0.32752709918551975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,10240,0.42740800645616317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,51200,3.2116302914089627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,7168,0.298239098654853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,8192,0.352967103322347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,7168,0.11066399680243598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,6144,0.24346934424506295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,7168,0.2854044437408447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,6144,0.09063288900587295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,5120,0.2060826619466146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,51200,2.012298583984375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,16384,0.735405339135064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,5120,0.2056737740834554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,4096,0.0675777792930603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,8192,0.11856799655490452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,3584,0.13154755698310003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,4096,0.16539733939700657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,3072,0.11326755417717828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,3584,0.1456800036960178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,3072,0.05300355619854397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,2560,0.0988960001203749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,3072,0.12710400422414145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,10240,0.4412684440612793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,2560,0.04663199848598904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,2560,0.10851644145117866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,2048,0.03996088769700792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,2048,0.08972799777984619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,1536,0.05923110908932156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,1536,0.03385333220163981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,1536,0.07172711027993096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,6144,0.24551023377312553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,2048,0.08616533544328477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,768,0.03401688734690348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,1024,0.053695999913745456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,1024,0.027894222074084814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,512,0.02716711163520813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,768,0.024554666545655992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,768,0.043531556924184166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,512,0.022457778453826904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,512,0.0358515547381507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,1024,0.040991110934151545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,256,0.020439111524158053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,4096,0.17370133929782441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,128,0.02017866737312741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,64,0.015167999598715039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,3584,0.05866666634877523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,256,0.019229332605997723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,64,0.029880887932247583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,32,0.02939288814862569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,256,0.03178044491344028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,5120,0.07817600170771281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,128,0.030100444952646892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,32,0.01587466730011834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,65536,0.9929724799262153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,128,0.01756000022093455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,51200,0.7654319869147407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,65536,2.135955598619249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,16384,0.21088000138600668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,51200,2.0472674899631076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,12288,0.15459556049770778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,16384,0.6196542316012913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,12288,0.46229510837131077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,10240,0.1304639975229899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,16384,0.5659191343519423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,10240,0.36369866795010036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,51200,2.9130853017171225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,8192,0.10796711179945205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,7168,0.23957512113783094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,8192,0.2907466623518202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,7168,0.09724444150924683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,6144,0.2366417778862847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,7168,0.31249245007832843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,6144,0.08451910813649495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,6144,0.21809866693284777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,5120,0.07255644268459745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,8192,0.27956265873379177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,5120,0.18524355358547637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,4096,0.05901066462198893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,65536,3.6812044779459634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,4096,0.14628267288208008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,3072,0.10559199915991889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,3584,0.12971288628048366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,10240,0.43139200740390354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,2560,0.08819378084606594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,3072,0.04740800129042732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,3072,0.1129831075668335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,2048,0.07421244515313043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,2560,0.04159733321931627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,2560,0.09684089157316421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,2048,0.03553955422507392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,2048,0.08037333356009589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,3584,0.05328266819318136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,1024,0.04279111160172356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,1536,0.029881778690550063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,5120,0.1941066715452406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,768,0.0317573348681132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,1024,0.024950222836600408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,4096,0.1256764464908176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,1024,0.04849688874350654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,512,0.023909333679411147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,768,0.022488888767030504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,768,0.039928889936871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,3584,0.11783111095428467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,12288,0.49281687206692165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,512,0.03212177753448486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,128,0.015623110863897534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,256,0.018421333697107103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,256,0.028119110398822363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,64,0.013726222018400827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,128,0.018399111098713346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,1536,0.0641511082649231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,128,0.026927999324268762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,64,0.02646222213904063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,32,0.026239110363854304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,256,0.01817599932352702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,512,0.020336000455750358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,1536,0.0574222207069397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,32,0.014123555686738757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,65536,0.8653448952568902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,51200,0.6436595386928982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,65536,1.8452027638753254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,16384,0.22556445333692762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,51200,1.5648568471272786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,16384,0.5368160141838921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,12288,0.4668586519029405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,12288,0.15091466903686523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,10240,0.2996782196892632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,10240,0.1132871045006646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,16384,0.5134097735087076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,8192,0.2425839900970459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,65536,2.8553369310167103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,8192,0.09498577647738987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,7168,0.21072799629635283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,8192,0.2515430980258518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,7168,0.08472977744208442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,6144,0.18625333574083117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,7168,0.23521600829230416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,6144,0.07331022289064196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,6144,0.19035822815365264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,12288,0.40201332834031844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,4096,0.11609866884019639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,5120,0.16091288460625544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,10240,0.4454808764987522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,51200,2.328364478217231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,4096,0.12922844621870253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,3584,0.04714133342107137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,3584,0.1130888859430949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,5120,0.1464960045284695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,3072,0.0415653321478102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,2560,0.07788088586595324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,5120,0.06344355477227105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,2048,0.067793779902988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,2560,0.03683288892110189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,1536,0.050352000527911715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,3584,0.10502044359842937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,2048,0.07046488920847575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,1536,0.026731555660565693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,1024,0.03512888815667894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,1536,0.05616622500949436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,768,0.026350221700138513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,1024,0.022094221578703985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,1024,0.042633778519100614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,3072,0.09924088584052192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,512,0.020336889558368258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,768,0.0347013341055976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,256,0.015375110838148328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,2560,0.08526311318079631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,4096,0.052505777941809766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,128,0.013448889056841532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,256,0.016864000095261466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,256,0.024526221884621516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,64,0.01481599940194024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,128,0.023718222975730896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,32,0.015509333875444202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,768,0.019924443629052904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,64,0.02335111134582096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,512,0.028428445259730022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,2048,0.03086844417783949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,3072,0.09522133403354222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,128,0.016699555847379897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,65536,0.7133786413404676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,512,0.01812533371978336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,32,0.02335111134582096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,65536,1.731244405110677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,16384,0.45487557517157656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,16384,0.1555751164754232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,16384,0.4525555504692926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,12288,0.11702311038970947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,10240,0.2842080063290066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,51200,1.8409315745035808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,10240,0.1008675562010871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,51200,0.5439448886447483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,8192,0.22748888863457573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,10240,0.2696479956309001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,8192,0.08112444480260213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,12288,0.30071110195583767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,7168,0.0718924469417996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,6144,0.14318133725060359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,12288,0.4141591125064426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,6144,0.06321333514319526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,6144,0.16302043861813015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,51200,1.3840648863050673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,5120,0.05451377895143297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,7168,0.16892888810899523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,5120,0.13743110497792563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,8192,0.21702578332689074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,3584,0.08833066622416179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,4096,0.0445128877957662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,4096,0.11031644211875068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,3072,0.08140089114507039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,65536,2.5146310594346786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,3584,0.09815822045008342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,2560,0.06369688775804308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,5120,0.14121688736809626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,3072,0.08563555611504449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,2048,0.058507555060916476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,2560,0.03196000059445699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,2560,0.07323022021187676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,2048,0.026994667119450037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,4096,0.1074062254693773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,2048,0.061556445227728955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,1024,0.030280888080596924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,1536,0.022442667020691767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,1536,0.04896266592873467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,1024,0.01908533275127411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,768,0.02163999941613939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,1024,0.03713333275583055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,512,0.016935111747847665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,768,0.017681777477264404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,3072,0.035298665364583336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,768,0.030121776792738173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,512,0.01626133256488376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,512,0.02486755616135067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,1536,0.04240177737341987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,128,0.011404444773991903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,256,0.021317332983016968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,64,0.009972444838947719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,128,0.01513155632548862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,32,0.010013333625263637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,7168,0.18887021806504992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,3584,0.03955199983384874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,64,0.02033955521053738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,32,0.0203448881705602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,256,0.013646221823162503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,256,0.015218666858143277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,65536,0.5675662358601888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,128,0.020615110794703167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,16384,0.3228488763173421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,51200,0.9233058293660482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,65536,1.4935483932495117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,16384,0.1275413301255968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,16384,0.4226071039835612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,10240,0.2120586633682251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,12288,0.26851288477579754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,65536,1.9197983211941188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,10240,0.09262488947974311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,10240,0.22702577379014757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,8192,0.06894399722417195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,7168,0.14855644438001844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,8192,0.18907911247677275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,7168,0.06153955724504259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,7168,0.16130488448672825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,12288,0.13463911745283338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,6144,0.05405866768625048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,6144,0.13772888978322348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,8192,0.1613911125395033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,4096,0.08143999841478136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,5120,0.11608266830444336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,4096,0.038204444779290095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,3584,0.06870844629075792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,5120,0.10153600242402817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,51200,1.430397351582845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,3072,0.0626844432618883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,3584,0.03347999850908915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,51200,0.45373953713311094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,3072,0.029902223083708022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,2560,0.05448977814780342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,3072,0.07180800040562947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,2048,0.041518221298853554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,2560,0.027217777238951787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,2560,0.06236088938183255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,1536,0.0337119996547699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,2048,0.05171822177039253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,3584,0.08255110846625434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,1536,0.02017688916789161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,1536,0.04196444484922621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,1024,0.01719911065366533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,12288,0.25635290145874023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,1024,0.031031999323103163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,768,0.01585244470172458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,2048,0.02335111134582096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,5120,0.04549422197871738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,512,0.02118844456142849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,256,0.011744889120260874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,512,0.014478221535682678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,6144,0.12294844786326091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,768,0.019814221395386588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,128,0.009712888962692684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,512,0.014510222607188754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,128,0.01756533318095737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,64,0.008932444784376357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,32,0.00904444439543618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,1024,0.02509422269132402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,64,0.01719377769364251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,32,0.017272000511487324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,4096,0.09211822350819905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,256,0.018242667118708294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,768,0.02544533378548092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,65536,0.48798312081231016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,65536,1.144026650322808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,51200,0.7704470952351888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,16384,0.2792782253689236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,51200,0.3897839917076959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,16384,0.12844178411695692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,12288,0.20044444666968453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,65536,1.6376027001274958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,16384,0.40568889511956113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,51200,1.200448883904351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,12288,0.2454684575398763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,128,0.01368444412946701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,10240,0.07719466421339247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,8192,0.06175644530190361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,10240,0.20606399907006157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,8192,0.1613395611445109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,7168,0.05521777934498257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,7168,0.14168800248040095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,6144,0.11750045087602402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,12288,0.09308711025449966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,6144,0.04805066519313388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,6144,0.12206933233473037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,8192,0.1691866715749105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,4096,0.07918400234646268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,7168,0.13974844084845647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,5120,0.10383199983172947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,3584,0.06643911202748616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,4096,0.08425066868464152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,256,0.013976888524161445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,3072,0.0558462209171719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,3584,0.030927111705144245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,5120,0.09611466858122085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,3072,0.028201778729756672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,2560,0.04738400048679776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,3072,0.06539199749628703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,2048,0.03976533479160733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,5120,0.040976001156700984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,2560,0.02533866630660163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,1536,0.027907556957668726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,2048,0.021744888689782884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,2048,0.046893331739637584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,1024,0.02130222154988183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,1536,0.037921776374181114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,10240,0.1835191117392646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,1024,0.0282008879714542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,768,0.015072888798183866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,512,0.013350221845838757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,2560,0.055663108825683594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,4096,0.03426488902833726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,256,0.010195555786291758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,512,0.01391377713945177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,512,0.01920355525281694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,128,0.009009777671760982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,3584,0.07507999738057454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,256,0.01680888897842831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,64,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,128,0.01581333412064446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,768,0.023231999741660223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,32,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,64,0.015604444675975375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,32,0.015855110353893705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,1536,0.018553778529167175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,256,0.013169777890046438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,768,0.017172444197866652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,128,0.013056000073750814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,65536,1.0271387100219727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,51200,0.3397226598527696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,51200,0.7571395768059624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,16384,0.10417866706848145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,16384,0.309808890024821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,51200,1.022473759121365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,1024,0.015850666496488783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,10240,0.1803119977315267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,65536,0.4285706679026286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,10240,0.07187377744250827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,8192,0.13654222753312853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,16384,0.2477511035071479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,7168,0.11906222502390544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,12288,0.20757155948215059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,65536,1.376612451341417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,8192,0.14377244313557944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,6144,0.1045715543958876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,7168,0.04983555608325534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,12288,0.08053600125842623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,5120,0.08660799927181667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,6144,0.04320355587535434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,12288,0.21044445037841797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,4096,0.0653582215309143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,5120,0.036943111154768206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,5120,0.09243110815684001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,4096,0.03090933296415541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,4096,0.07528533538182576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,8192,0.05535466803444756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,3584,0.027847111225128174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,3584,0.06638044118881226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,3072,0.025571554899215698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,7168,0.12622222635481092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,2560,0.023153778579499986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,2560,0.0491226679748959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,2048,0.03384888834423489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,3584,0.05494311120775011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,1536,0.026405332816971674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,2048,0.04155644443300035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,1536,0.017802667286660936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,1536,0.033675554725858904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,1024,0.019221333993805777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,3072,0.05723466475804647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,768,0.01589777734544542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,6144,0.10902221997578938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,10240,0.17833333545260957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,768,0.020976889464590285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,3072,0.04594666759173075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,2560,0.04110222061475118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,512,0.017458667357762653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,256,0.01072799993885888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,512,0.013146667016877068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,1024,0.015271999769740634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,128,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,256,0.015042665931913586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,768,0.014383110735151501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,2048,0.020589333441522386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,64,0.007355555891990662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,128,0.014304889572991265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,32,0.007648000286685095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,64,0.014156444205178155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,32,0.014702222413486905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,256,0.012455110748608908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,65536,0.8899786207411023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,512,0.012137778103351593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,51200,0.577007982465956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,16384,0.22300799687703451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,65536,1.143530633714464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,16384,0.09223200215233697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,12288,0.16914844512939453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,1024,0.025053333905008104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,51200,0.8783582051595052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,10240,0.13037599457634821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,128,0.012537777423858643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,12288,0.18683200412326387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,10240,0.062237335575951465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,10240,0.1617013348473443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,8192,0.04882133338186476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,7168,0.0938533345858256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,7168,0.043613334496816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,16384,0.25180088149176705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,7168,0.10951199796464707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,12288,0.07716888851589628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,5120,0.06602311134338379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,6144,0.09522933430141872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,8192,0.10853155453999837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,4096,0.05455733338991801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,5120,0.03278311093648275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,5120,0.08142577939563327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,8192,0.1247404416402181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,3584,0.05041511191262139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,4096,0.02788089050187005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,4096,0.06541244188944499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,3584,0.02529688841766781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,6144,0.08172621991899279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,3584,0.05822844637764824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,65536,0.37909510400560165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,3072,0.02273422148492601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,2048,0.02868355645073785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,2560,0.020572443803151447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,2560,0.044312887721591525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,2048,0.018198221921920776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,1536,0.023225777679019507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,2048,0.03720266620318095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,51200,0.30051199595133465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,1024,0.017308443784713745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,1536,0.0158924443854226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,768,0.014682667122946845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,3072,0.03847733471128676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,1024,0.022289777795473736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,512,0.011768889096048145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,2560,0.033324443631702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,768,0.018570666511853535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,256,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,512,0.015642666154437594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,256,0.011696889168686338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,256,0.013455111119482251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,128,0.007294221884674496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,1024,0.014025777578353882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,64,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,128,0.01277955538696713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,768,0.013387555877367655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,6144,0.037978665696250066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,64,0.01276533305644989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,512,0.012416889270146688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,32,0.012788444757461548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,65536,0.3231297863854302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,65536,0.6329449017842611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,51200,0.4370017846425374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,16384,0.16610666116078696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,32,0.006608888920810487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,51200,0.2853013409508599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,16384,0.07959377765655518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,12288,0.12182044982910156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,16384,0.21144355667961967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,12288,0.06785155667199029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,10240,0.09981244140201145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,12288,0.15757155418395996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,1536,0.029584887954923842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,8192,0.09248266617457072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,10240,0.1356053352355957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,8192,0.04316622349951002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,7168,0.07816621992323133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,3072,0.05072355601522657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,6144,0.06715199682447645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,7168,0.03759288787841797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,51200,0.7095893224080404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,5120,0.05610311031341553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,6144,0.08205866813659668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,128,0.011772444678677453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,5120,0.028697778781255085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,4096,0.04489599996142917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,65536,0.9049733479817709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,8192,0.10830133491092259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,3584,0.03708088729116652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,4096,0.024392889605628118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,7168,0.09432711203893025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,3584,0.02239822182390425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,6144,0.0324906673696306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,3584,0.05009422368473477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,2560,0.029581334855821397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,3072,0.020406223005718656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,3072,0.04474844535191854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,5120,0.06939111153284709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,2560,0.018391110830836825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,2560,0.03860266672240363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,1536,0.019525332583321463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,2048,0.01667555504375034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,4096,0.05602577659818861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,1024,0.014191110928853353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,1536,0.02630133264594608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,1024,0.01313777764638265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,10240,0.054916444751951426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,768,0.01204533295498954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,2048,0.024519999821980793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,768,0.012378666963842181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,512,0.009732444253232744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,768,0.016751110553741455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,256,0.007122666471534305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,2048,0.03243111239539252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,128,0.006274666637182236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,1536,0.015028445257080926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,256,0.011038222246699862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,3072,0.03461422191725837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,64,0.005635555419656966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,128,0.010870222416188983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,128,0.011416888899273343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,64,0.011065777805116443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,32,0.011372444530328115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,65536,0.509951114654541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,512,0.011409777734014722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,65536,0.2615413400861952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,256,0.011960888902346293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,51200,0.3963075478871663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,32,0.0063164445261160536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,51200,0.20708000659942627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,12288,0.09941689173380534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,51200,0.5430266592237685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,16384,0.2283102141486274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,12288,0.05838933255937365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,10240,0.07974488867653741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,12288,0.1299991077846951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,512,0.013978666729397245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,10240,0.05148177676730686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,10240,0.11001600159539117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,65536,0.6880204412672254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,16384,0.12476444244384766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,8192,0.08937422434488933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,6144,0.04694933361477322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,7168,0.033174223370022245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,7168,0.07767911089791192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,5120,0.03949600127008226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,6144,0.06782044304741754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,8192,0.06623378064897326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,5120,0.02499733368555705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,4096,0.031678222947650485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,8192,0.03723466727468703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,4096,0.021664000219768945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,1024,0.01979022224744161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,4096,0.04635555545488993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,16384,0.07097955544789632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,3584,0.019148444135983784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,3584,0.0414266652531094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,6144,0.029426667425367568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,3072,0.03670755690998501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,2560,0.016535111599498324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,5120,0.05667466587490506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,7168,0.05387733379999796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,3584,0.028769777880774602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,1536,0.01629866659641266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,3072,0.02674755619631873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,1536,0.013167111410035027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,2560,0.022686221533351477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,1536,0.021948445174429152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,1024,0.01178311142656538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,768,0.009359111388524374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,2048,0.01516888870133294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,768,0.011074666347768573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,512,0.008267555799749162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,3072,0.01756355497572157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,512,0.010435555544164447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,256,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,2560,0.03175466590457492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,256,0.010032888915803697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,256,0.010144000252087912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,768,0.014118222726715935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,128,0.010025777750545079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,64,0.005283555636803309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,1024,0.011730666789743634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,32,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,512,0.012091555529170565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,1024,0.01678222252262963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,32,0.009717333647939894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,128,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,2048,0.027168888184759352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,128,0.009704888694816166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,65536,0.5519626405504015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,64,0.009376000199053023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,2048,0.018780445059140522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,51200,0.16336711247762045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,16384,0.1000559992260403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,51200,0.4263128969404433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,12288,0.07723822196324666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,65536,0.38513776991102433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,16384,0.13725688722398546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,10240,0.06184444162580702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,12288,0.10482577482859294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,51200,0.29711998833550346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,10240,0.039248890346950956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,8192,0.04979377653863695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,10240,0.08778666787677342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,8192,0.07095555464426677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,16384,0.055492444170845874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,6144,0.03456088900566101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,7168,0.06236622068617079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,12288,0.04438044296370613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,5120,0.02799644403987461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,65536,0.204346670044793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,6144,0.05459466576576233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,4096,0.022692445251676772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,5120,0.045863999260796436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,3584,0.02088266611099243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,4096,0.038021332687801786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,7168,0.043324444029066295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,3072,0.017550221747822232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,3584,0.016132444143295288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,7168,0.02607822252644433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,3584,0.03399822115898132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,2560,0.01533422205183241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,3072,0.014819555812411837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,3072,0.030231111579471167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,6144,0.022499556342760723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,2560,0.013792888985739814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,1536,0.011056889262464313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,2048,0.012430222498046027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,2048,0.02221600049071842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,1024,0.008958222137557136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,1536,0.011536888778209686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,1536,0.01827555563714769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,768,0.007294221884674496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,1024,0.010604444477293225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,1024,0.014025777578353882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,768,0.010048000348938836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,768,0.01202933324707879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,2048,0.013207111093733044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,256,0.0053004444473319584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,512,0.009550222092204625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,512,0.010382222632567087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,5120,0.019938665959570143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,256,0.00925333301226298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,256,0.00871288859181934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,4096,0.017342221405771043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,64,0.004223111189073986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,128,0.00907733374171787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,64,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,512,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,32,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,65536,0.2562577724456787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,2560,0.026123555170165166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,65536,0.20059200127919516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,128,0.004563555535342959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,51200,0.19209688239627412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,16384,0.07181421915690105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,65536,0.4590826564364963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,51200,0.15944711367289224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,12288,0.054758220911026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,16384,0.055063999361462064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,51200,0.3659031126234267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,16384,0.11850666999816895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,12288,0.04354044463899401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,8192,0.03642488850487603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,12288,0.08949777815077041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,10240,0.039749334255854286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,7168,0.0302782224284278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,8192,0.028129776318868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,8192,0.0609475572903951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,6144,0.026870222555266485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,7168,0.054139554500579834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,32,0.004568888909286923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,6144,0.02237244447072347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,5120,0.022267555197079975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,10240,0.044975111881891884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,4096,0.017656889226701524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,5120,0.019732443822754752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,5120,0.03971822063128153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,3584,0.016601777738995023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,4096,0.03286666671435038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,7168,0.025397333833906386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,3584,0.02947910957866245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,3072,0.014702222413486905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,128,0.008023999631404877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,3072,0.01499288943078783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,2560,0.01276177747382058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,3072,0.02613688839806451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,2560,0.013807999591032663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,10240,0.07602222098244561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,3584,0.015963556038008798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,2048,0.0192586663696501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,1536,0.009167110754383935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,6144,0.04681155416700575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,1536,0.01148000028398302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,1024,0.0075342224703894714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,4096,0.017260443833139207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,1024,0.010728889041476779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,1024,0.012430222498046027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,768,0.006612444503439798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,8192,0.02844266759024726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,512,0.005607999861240387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,2048,0.012507555385430654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,768,0.01071111112833023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,256,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,512,0.009581333233250512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,256,0.009354666703277165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,2048,0.010992888775136737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,128,0.009103111094898647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,64,0.0042275554604000514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,768,0.010031999813185798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,32,0.004259555704063839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,64,0.007125332951545715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,1536,0.01590133375591702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,65536,0.19997066921657988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,32,0.007296000089910295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,2560,0.022712000542216833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,51200,0.16369244787428114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,65536,0.154095106654697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,128,0.007484444313579136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,16384,0.052335109975602895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,65536,0.4093457857767741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,12288,0.039968000517951116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,51200,0.3158622317843967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,16384,0.10542222526338364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,512,0.009695111049546136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,10240,0.03202666507826911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,12288,0.08135910828908284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,10240,0.030900445249345567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,8192,0.02570844524436527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,256,0.007887110941939885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,10240,0.06812355253431532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,7168,0.021529777182473078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,8192,0.021524444222450256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,6144,0.019855111837387085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,7168,0.019632889164818656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,51200,0.12386311425103082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,7168,0.04804799954096476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,5120,0.016747555798954435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,6144,0.04194666610823738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,5120,0.01555022266176012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,12288,0.03502488798565335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,5120,0.035863998863432146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,3584,0.012877333495352002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,4096,0.013804444836245643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,4096,0.029495110114415485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,3072,0.011575111084514193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,3584,0.026550221774313185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,8192,0.05431644452942742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,2560,0.009850666754775578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,3072,0.012190221912331052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,3072,0.023753777146339417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,2048,0.008672888908121321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,2560,0.01166044423977534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,2560,0.02071822186311086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,1536,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,128,0.004591111093759537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,1536,0.010194666683673859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,3584,0.013125333521101209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,1024,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,1024,0.009746666583749983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,768,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,4096,0.014058666096793281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,2048,0.01088355564408832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,512,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,768,0.009163555171754625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,512,0.00869599978129069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,256,0.003987555702527364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,16384,0.0432586669921875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,256,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,128,0.003547555456558863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,256,0.008341333104504479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,768,0.00999022192425198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,128,0.00832799987660514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,128,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,32,0.003555555724435382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,1536,0.014816888504558139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,64,0.00685066646999783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,65536,0.11661688486735027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,32,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,6144,0.017291555802027386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,51200,0.1024506688117981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,65536,0.15351288848453098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,65536,0.32833777533637154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,51200,0.12290310859680176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,64,0.0032826666202810076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,51200,0.2551040119594998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,12288,0.0278479986720615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,16384,0.04316177633073595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,512,0.00869066682126787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,10240,0.02102488941616482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,12288,0.06471111377080281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,1024,0.01146399974822998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,8192,0.017589333984586928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,10240,0.03038133184115092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,10240,0.05457955598831177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,16384,0.039459556341171265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,8192,0.021253334151373968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,2048,0.017679111825095285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,8192,0.04424622323777941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,6144,0.014194667339324951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,7168,0.01870844430393643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,5120,0.0125937776433097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,6144,0.017581333716710407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,6144,0.034734222624037005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,5120,0.01568711135122511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,5120,0.02950577934583028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,4096,0.013804444836245643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,3584,0.010052444206343757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,7168,0.015597333510716757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,3584,0.012893333203262754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,3072,0.008848889006508721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,3584,0.02195555633968777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,7168,0.03956266575389438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,2560,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,3072,0.012451555165979596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,12288,0.03489155570665995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,4096,0.010038221875826517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,2560,0.011530666715568967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,2560,0.017267554998397827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,2048,0.010860444770918952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,2048,0.014874666929244995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,4096,0.024151111642519634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,1536,0.01253600004646513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,1024,0.0094035557574696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,1024,0.010071111222108206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,768,0.004936000125275718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,768,0.00908799966176351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,512,0.004260444392760594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,1536,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,512,0.00868444475862715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,1536,0.010335110955768162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,256,0.0038862224254343244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,256,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,128,0.003560888684458203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,3072,0.01975200076897939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,768,0.008767999708652496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,64,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,512,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,32,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,16384,0.08610133330027263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,32,0.005648889061477449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,65536,0.08743466933568318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,2048,0.0070053330726093715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,128,0.005991111199061076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,51200,0.07189777824613783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,128,0.00831822223133511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,65536,0.15368355645073786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,16384,0.029066665305031672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,65536,0.2850080066257053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,1024,0.005580444302823808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,12288,0.02243288854757945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,51200,0.22270221180386016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,16384,0.07464710871378581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,12288,0.03537777728504605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,12288,0.058000889089372426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,8192,0.013481777575280933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,10240,0.030806223551432293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,10240,0.048093332184685596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,7168,0.012127111355463663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,8192,0.039270222187042236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,51200,0.12354044119517009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,7168,0.02017688916789161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,64,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,16384,0.04423022270202637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,5120,0.010037333601050908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,6144,0.018236445056067575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,6144,0.03013155526585049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,5120,0.01620444489849938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,5120,0.025919111238585577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,8192,0.0225911107328203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,4096,0.014488000008794995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,4096,0.02164799968401591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,3584,0.013497778111033969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,7168,0.03437866767247518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,3072,0.007282666862010956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,10240,0.019253333409627277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,256,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,2560,0.006808888994985157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,3072,0.0177208897140291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,4096,0.0086666668454806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,2048,0.00590133335855272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,2560,0.015805333852767944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,1536,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,3584,0.007663110891977946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,2048,0.011065777805116443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,1536,0.01036622209681405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,1536,0.011413333316644033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,3072,0.012777778009573618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,1024,0.009352888498041365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,768,0.004001777619123459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,2560,0.011777777638700275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,768,0.009053332938088311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,768,0.008028444316652086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,512,0.0038444445365005066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,6144,0.011169777976142036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,3584,0.019727999965349834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,512,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,256,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,256,0.005962666538026597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,128,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,64,0.0028693332440323303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,128,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,2048,0.013549333645237817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,32,0.0032311110860771606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,65536,0.08973066674338447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,64,0.00526577772365676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,32,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,51200,0.07275733020570543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,16384,0.026439110438028973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,128,0.0032311110860771606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,12288,0.01650311052799225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,51200,0.2287671036190457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,16384,0.07483288976881239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,10240,0.01292711082432005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,8192,0.012086222569147745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,12288,0.05683288971583048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,10240,0.047982222504085965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,8192,0.039112889104419284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,6144,0.010074666804737514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,7168,0.034690668185551964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,1024,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,6144,0.029951112137900457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,4096,0.007985778152942657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,5120,0.025423111187087163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,3584,0.007282666862010956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,4096,0.021543110410372417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,3072,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,3584,0.019369777705934312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,2560,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,65536,0.28412800365024143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,2048,0.0052915554907586836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,2560,0.01551822159025404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,1536,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,2048,0.013456888496875763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,1024,0.003899555653333664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,1024,0.0042071110672420925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,768,0.003879111260175705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,1024,0.009305777649084726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,512,0.003559999995761447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,5120,0.009018667042255402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,256,0.003208000005947219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,512,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,128,0.0028702221396896574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,256,0.005791999813583162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,64,0.0028515555378463534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,512,0.008665777742862701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,32,0.002847111059559716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,64,0.00526577772365676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,65536,0.08743644422954983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,3072,0.01760444376203749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,51200,0.06940533055199517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,65536,0.2842000060611301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,7168,0.011775111158688864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,128,0.0052791109515561005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,12288,0.013457777599493662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,16384,0.07478666967815824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,768,0.008034666379292807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,10240,0.011786667009194693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,12288,0.05624622106552124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,10240,0.04784799946679009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,8192,0.03870044483078851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,256,0.0032542221662071017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,6144,0.009405333134863112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,7168,0.034240000777774386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,5120,0.011753777662913004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,32,0.0052373334765434265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,6144,0.03012088934580485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,4096,0.00998577806684706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,8192,0.01070844464831882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,5120,0.025660443637106154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,3072,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,4096,0.021706667211320665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,3584,0.020003555549515616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,16384,0.025427555044492085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,2048,0.005250666704442766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,3072,0.01789688898457421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,1536,0.004701333327425851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,2048,0.013805333111021252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,1024,0.003932444585694207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,3584,0.009027555584907532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,768,0.0038933331767717996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,1024,0.009364444348547194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,7168,0.010020444790522257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,512,0.0032684444967243406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,768,0.008015111088752747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,1536,0.011404444773991903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,256,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,128,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,512,0.007089777953094906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,64,0.0028488888508743713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,256,0.005681777579916849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,32,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,128,0.005393777870469623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,1536,0.011526222030321756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,64,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,32,0.0052826665341854095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,2560,0.007585778004593319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,51200,0.22853512234157988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,2560,0.01590044465329912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,16384,1.527672873602973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,16384,4.878241644965278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,16384,5.975507524278428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,51200,6.420800103081597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,16384,4.888979593912761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,12288,3.6675813462999134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,12288,3.789946662055122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,12288,1.1389688915676541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,10240,3.107028537326389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,10240,0.9527173572116427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,10240,3.168850580851237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,12288,3.5533582899305554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,8192,0.7619199752807617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,10240,2.9001831478542752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,8192,2.6123936971028647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,51200,14.837131076388888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,8192,2.3530871073404946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,8192,2.4903216891818576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,7168,2.15500619676378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,7168,2.141783184475369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,7168,0.6869209077623156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,6144,1.7824925316704645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,6144,0.5970444679260254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,7168,2.046688927544488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,5120,0.49831379784478086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,6144,1.7866872151692708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,5120,1.7929083506266277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,6144,1.7955057356092665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,5120,1.4933564927842882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,4096,0.415775113635593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,4096,1.2577963935004342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,5120,1.4611262215508356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,4096,1.190267562866211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,3584,1.0402658250596788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,4096,1.1717821756998699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,3072,0.8259341981675891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,3584,1.0415484110514324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,3072,0.9030479855007596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,3584,1.0169955359564888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,3072,0.3323662281036377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,3584,0.37465333938598633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,3072,0.8838479783799914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,2560,0.2904711034562853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,2048,0.5882186889648438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,2560,0.7126524183485242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,2048,0.6287991205851237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,2048,0.2422888808780246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,2560,0.6769599914550781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,2048,0.596544000837538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,51200,22.714545355902775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,1536,0.19404355684916177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,1024,0.33131111992730033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,2560,0.7468746503194174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,1536,0.4622204568650987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,1024,0.1609626611073812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,768,0.25010845396253795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,1024,0.3332515557607015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,1536,0.43406221601698136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,51200,18.87743631998698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,768,0.29859288533528644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,768,0.28112533357408315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,1536,0.48587465286254883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,512,0.1210800011952718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,512,0.24682667520311144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,1024,0.37668800354003906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,128,0.10297422276602851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,256,0.10564177566104466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,256,0.17611644003126356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,768,0.13658311631944445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,512,0.19418133629692924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,64,0.08556444115108913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,128,0.20914222134484184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,128,0.13458755281236437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,256,0.13695732752482095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,64,0.21388710869683158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,32,0.216538667678833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,256,0.22244887881808809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,512,0.34451643625895184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,128,0.10370222065183853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,32,0.08856888612111409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,51200,4.914061228434245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,65536,6.2380320231119795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,16384,4.131204393174913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,51200,11.753241644965279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,16384,4.451216803656684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,16384,1.2647778193155925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,65536,15.340801662868925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,16384,3.8388159010145397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,12288,3.1925616794162326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,12288,0.9502418306138781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,51200,13.129809061686197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,51200,18.14406500922309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,12288,3.107305738661024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,12288,2.752745734320747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,10240,0.7376160091824002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,10240,2.4197794596354165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,8192,1.9598301781548395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,8192,0.5943928824530708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,10240,2.556004418267144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,8192,1.9216382768419054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,10240,2.2716195848253036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,7168,0.5298666424221462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,7168,1.7013253106011286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,8192,1.7745449278089735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,7168,1.7100640402899847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,65536,23.375688340928818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,6144,1.389327155219184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,7168,1.598354657491048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,6144,0.4585253397623698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,65536,18.815277099609375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,6144,1.4117991129557292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,5120,1.3938408957587347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,4096,0.9342044194539388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,5120,1.20215818617079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,6144,1.3339804543389215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,4096,0.3556239869859483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,5120,0.3924666775597467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,3584,0.8436924616495768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,3584,0.29534843232896596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,4096,0.861030207739936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,4096,0.9270195431179471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,3584,0.7527848879496256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,3072,0.2539173232184516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,5120,1.1470097435845268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,2560,0.6011315451727973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,3072,0.7451110945807563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,2560,0.23127910825941297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,2560,0.6016577614678277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,3584,0.8005653487311469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,2560,0.5427369011773003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,2048,0.18993866443634033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,1536,0.3712826569875081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,2048,0.4613662295871311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,1536,0.15620088577270508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,3072,0.6629288991292318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,1024,0.25798487663269043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,2048,0.547884464263916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,1024,0.12432889143625896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,1024,0.28710754712422687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,3072,0.679532421959771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,768,0.10782844490475124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,768,0.22673156526353624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,1536,0.3921822177039252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,768,0.2172773281733195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,1536,0.3538373311360677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,512,0.097126219007704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,512,0.19415999783409965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,768,0.1960657835006714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,512,0.1766479942533705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,1024,0.26578132311503094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,2048,0.4881546762254503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,256,0.08480177985297309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,256,0.17538932959238687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,256,0.1396968894534641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,64,0.0713333355055915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,128,0.08369333214230007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,128,0.1649671130710178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,512,0.14906577269236246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,64,0.16851911279890272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,256,0.10274844699435765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,32,0.17052177588144937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,128,0.08146578073501587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,32,0.07238755623499553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,128,0.10701600048277114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,65536,1.9351075490315754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,51200,1.4200959735446508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,16384,1.3836515214708116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,16384,1.4861324098375108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,65536,5.170425838894315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,16384,0.3830808798472087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,12288,0.9148622088962131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,51200,4.11643812391493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,51200,3.830705854627821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,65536,5.310835520426433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,16384,1.2556053797403972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,10240,0.7680471208360461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,12288,0.8288737932840983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,51200,6.1811616685655375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,12288,1.0068960189819336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,10240,0.29132623142666286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,65536,7.461927625868056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,10240,0.9920408460828992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,12288,0.2968488799201118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,8192,0.6121110916137695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,8192,0.5303822093539768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,7168,0.53657775455051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,7168,0.17921955055660674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,7168,0.5351929134792751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,6144,0.4289813306596544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,7168,0.4840124448140462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,8192,0.6461742189195421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,6144,0.45722044838799375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,8192,0.19955200619167754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,5120,0.3977359930674235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,6144,0.16463732719421387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,6144,0.41407733493381077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,5120,0.13464532958136663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,4096,0.2936017778184679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,5120,0.3863982094658746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,5120,0.3389742109510634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,3584,0.28099110391404897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,4096,0.30710045496622723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,4096,0.28338223033481175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,3584,0.269287109375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,3072,0.21527466509077284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,3584,0.2443173461490207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,3072,0.08916355503929986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,3072,0.23349067899915907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,2560,0.19238844182756212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,3072,0.22003466553158232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,2048,0.15342133575015598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,2560,0.20021866427527535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,2560,0.186191995938619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,10240,0.6805342038472494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,3584,0.11628088686201309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,1536,0.1166657739215427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,2048,0.15554489029778376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,1536,0.05537955628501045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,1536,0.1301520003212823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,1536,0.12343645095825195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,4096,0.11179110738966201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,768,0.06524799929724799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,2048,0.06657155354817708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,768,0.07910399966769747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,1024,0.08301333586374919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,768,0.07754400041368273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,768,0.042429334587521024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,1024,0.09778222110536362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,512,0.06735110945171781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,1024,0.09460444582833184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,512,0.06519644127951728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,512,0.03868710994720459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,2048,0.16361599498324925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,128,0.027887112564510767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,256,0.03432088759210374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,256,0.06056977642907036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,256,0.05095555716090732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,64,0.02510044475396474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,128,0.05695911248524984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,32,0.026123555170165166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,128,0.04100622071160211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,2560,0.07743910948435466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,256,0.03346755438380771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,32,0.05870844258202446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,512,0.053504887554380626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,128,0.03397511111365424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,1024,0.04569866591029697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,65536,1.4351209004720051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,64,0.05815644396675957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,65536,3.674119313557943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,16384,0.9428461922539605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,51200,2.9860276116265188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,16384,0.30131199624803334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,51200,1.0438790851169162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,16384,1.0238702562120225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,16384,0.9017395443386502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,12288,0.7198559972974988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,65536,3.960693359375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,12288,0.22478045357598198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,51200,4.897344801161024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,65536,5.876750098334418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,12288,0.7363191180759006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,12288,0.6294124391343858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,10240,0.20902933014763725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,8192,0.4916337860955133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,10240,0.6542755762736002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,8192,0.1546186606089274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,10240,0.5107306904262966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,51200,3.078151914808485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,8192,0.47676976521809894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,8192,0.4170080025990804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,6144,0.3597262170579698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,10240,0.6607822312249078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,7168,0.3589591185251872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,6144,0.35387288199530703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,5120,0.29067111015319824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,6144,0.31990665859646267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,5120,0.10691733492745294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,5120,0.2971537907918294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,4096,0.23079111840989855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,5120,0.2602915498945448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,7168,0.5341493288675944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,4096,0.08801866902245416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,7168,0.4127315680185954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,4096,0.23478221893310547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,4096,0.2131964498096042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,3584,0.07833155658509997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,3584,0.20454488860236275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,3072,0.17913333574930826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,3584,0.19239466720157197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,3072,0.068759110238817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,3072,0.17920622560713026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,3072,0.16622311539120144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,7168,0.13753422101338705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,2560,0.15595466560787624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,2560,0.060878223843044706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,2560,0.1420204506980048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,6144,0.12554755475785997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,2048,0.05225333240297106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,2048,0.12657154930962458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,2048,0.12036355336507161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,2560,0.15212266974978977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,1536,0.044248888889948525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,1536,0.10083199871910943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,1024,0.06292355722851224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,1536,0.09686933623419867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,768,0.05257066753175524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,3584,0.19085421827104357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,1024,0.07694844404856364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,1024,0.07399021916919284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,512,0.036841776635911726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,768,0.034676445855034724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,1536,0.09111289183298747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,768,0.06194400125079685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,512,0.03186577889654372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,512,0.05272000034650167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,512,0.05212000012397766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,256,0.04712533288531833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,256,0.0406275557147132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,128,0.044277333550983004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,768,0.061627553568945996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,128,0.028059555424584284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,256,0.027235555979940627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,32,0.01982311076588101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,64,0.045069333579805165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,128,0.02348533272743225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,32,0.04555111130078634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,1024,0.03722133239110311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,128,0.03289599882231818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,256,0.02836266822285122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,2048,0.1244844463136461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,65536,1.2364319695366752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,64,0.01991911066903008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,16384,0.8428257836235894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,65536,3.0898284912109375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,51200,0.9502977795071073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,51200,2.7201786041259766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,16384,0.3135297828250461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,16384,0.9564026726616753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,12288,0.6121884451972114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,16384,0.7438275549146863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,65536,3.351919174194336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,12288,0.6792302131652832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,65536,4.803023020426433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,12288,0.19943555196126303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,10240,0.16216088665856257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,51200,3.9542291429307728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,10240,0.5364329020182291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,10240,0.44145777490403915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,8192,0.13656977812449136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,7168,0.3461120128631592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,51200,2.577458699544271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,12288,0.5255644586351182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,7168,0.11863111125098334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,8192,0.4205617904663086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,7168,0.3442862298753526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,6144,0.2997617721557617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,8192,0.40155556466844344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,5120,0.2554391225179036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,6144,0.2951342264811198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,6144,0.27287199762132436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,5120,0.0890986654493544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,5120,0.2473377651638455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,7168,0.35019199053446454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,5120,0.23325777053833008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,4096,0.07566310962041219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,3584,0.1688542233573066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,4096,0.18371200561523438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,6144,0.10544000069300334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,8192,0.3744826581743028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,3072,0.14380978213416204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,3584,0.18264267179701063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,4096,0.19978755050235322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,3072,0.05936444467968411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,2560,0.12699466281467015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,4096,0.1964799960454305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,2560,0.130712005827162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,2560,0.05240355597601997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,2048,0.09452799956003825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,3584,0.1751697725719876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,2048,0.10768089029524062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,10240,0.4246719943152533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,2048,0.10325688785976833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,3072,0.14303021960788304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,1536,0.038266665405697294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,1536,0.08391111426883274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,1024,0.05394577648904589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,3584,0.06647821929719713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,2048,0.045569777488708496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,3072,0.15211110644870335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,1024,0.06611466407775879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,1536,0.08592622147666083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,768,0.053007112609015576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,2560,0.12359910541110569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,512,0.035550223456488714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,768,0.030545777744717065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,1024,0.06521333588494195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,1536,0.07233422332339816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,256,0.023373333944214716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,512,0.027923554182052612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,512,0.04549511273701986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,256,0.024647111694018047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,256,0.040285332335366145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,256,0.03506133291456435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,768,0.05383377936151293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,128,0.03779733180999756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,64,0.019519110520680744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,128,0.029159112109078303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,512,0.04483733243412442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,32,0.020024889045291476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,1024,0.03263911273744371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,64,0.038346668084462486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,32,0.03874844312667847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,128,0.02100444502300686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,128,0.02442577812406752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,768,0.045815110206604004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,65536,0.9340559641520182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,51200,0.7445919778611926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,16384,0.9447715547349719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,65536,2.7958123948838978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,16384,0.21022044287787545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,16384,0.8204213248358833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,16384,0.5967129071553549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,65536,2.6253634558783636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,12288,0.5165893236796061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,51200,2.2803253597683377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,12288,0.176200893190172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,51200,2.0262044270833335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,65536,4.286541408962674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,10240,0.4176906744639079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,12288,0.5007840262518989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,12288,0.42184000545077854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,8192,0.31902845700581867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,10240,0.4064951207902696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,8192,0.11066044701470269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,7168,0.2781439887152778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,8192,0.3792568842569987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,8192,0.2789733409881592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,7168,0.10872444179322983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,7168,0.2820764382680257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,7168,0.24743734465705025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,6144,0.08824888865152995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,10240,0.15060977141062418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,6144,0.22938222355312773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,10240,0.3559582233428955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,5120,0.20303644074334037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,5120,0.07322844531801012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,51200,3.0501342349582248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,5120,0.17920888794793022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,6144,0.23632889323764375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,4096,0.06119110849168566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,4096,0.1616284449895223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,4096,0.14899467097388372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,6144,0.2482586701711019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,5120,0.19993244277106392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,3584,0.14292800426483154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,3584,0.1347386704550849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,2560,0.09880177842246161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,3072,0.12550044059753418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,3072,0.11851378281911214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,3584,0.13273955716027155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,2560,0.10746310816870795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,2560,0.10063555505540635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,2048,0.08520533641179402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,3584,0.05657688776652018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,2048,0.03863999909824795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,2048,0.08944711420271133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,1536,0.06596266561084323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,2048,0.08471288945939805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,1024,0.042803555727005005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,3072,0.049815111690097384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,1536,0.0709822244114346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,1536,0.07026400168736775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,768,0.03306844499376085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,1024,0.05345511105325487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,1024,0.05328889025582207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,768,0.02423200011253357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,3072,0.11840089162190755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,512,0.02422222163942125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,1536,0.03233688738611009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,512,0.035466667678621076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,4096,0.16518933243221706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,512,0.0372924440436893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,1024,0.02607022225856781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,2560,0.04467733369933235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,256,0.031165334913465712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,256,0.020000888241661918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,128,0.016677333248986136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,256,0.029346668057971533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,768,0.04464000132348803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,64,0.016567111015319824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,128,0.02952444553375244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,32,0.016620443926917184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,256,0.01925155520439148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,64,0.029531555043326482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,32,0.02951822347111172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,768,0.04362755682733324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,128,0.020561777883105807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,128,0.02439555525779724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,512,0.02180533276663886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,65536,0.891076405843099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,65536,2.2992852528889975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,16384,0.46757157643636066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,51200,1.8411536746554906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,51200,0.7731128798590766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,16384,0.2190799978044298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,12288,0.4436195426517063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,16384,0.5365546544392904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,51200,2.540719985961914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,51200,1.907133314344618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,12288,0.46522845162285703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,12288,0.4730141957600911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,16384,0.6537093586391872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,10240,0.3969102170732286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,65536,2.544165293375651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,10240,0.3201875686645508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,8192,0.30738043785095215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,8192,0.10219999816682603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,7168,0.25619999567667645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,10240,0.33614667256673175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,8192,0.2542284329732259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,6144,0.21186044481065539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,10240,0.12348443932003444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,12288,0.14705422189500597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,8192,0.28589510917663574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,6144,0.21745777130126953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,6144,0.1953831116358439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,5120,0.18187199698554146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,7168,0.09150489171346028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,7168,0.22778399785359701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,5120,0.16628977987501356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,5120,0.0674640006489224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,4096,0.14338844352298313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,3584,0.1158026721742418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,4096,0.05654933055241903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,4096,0.1452888912624783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,4096,0.1376133362452189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,6144,0.07823289102978177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,3584,0.05165511038568285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,3584,0.12817511293623182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,3584,0.1211902168061998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,3072,0.04572088850869072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,2560,0.08304178052478366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,7168,0.24915289878845215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,3072,0.10707733366224502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,2048,0.07027910815344916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,2560,0.09596533245510525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,2560,0.0930044452349345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,5120,0.18470933702256945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,2048,0.034613334470325045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,3072,0.10602044396930271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,1536,0.058026664786868624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,2048,0.07707289192411634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,1024,0.038703110482957624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,1536,0.028593778610229492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,3072,0.11179555786980523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,1536,0.0633760028415256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,1536,0.06452000141143799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,1024,0.024720888998773363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,1024,0.047839999198913574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,1024,0.0491848886013031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,512,0.02163822286658817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,768,0.03900444507598877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,768,0.041484445333480835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,512,0.03171822097566392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,256,0.01711466742886437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,65536,3.5353465610080295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,768,0.027493332823117573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,256,0.027845333019892376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,128,0.015078221758206686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,256,0.02688088847531213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,2048,0.07967022392484877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,128,0.02641511129008399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,64,0.012876444392734103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,512,0.020399111840460036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,128,0.022844443718592327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,32,0.013194666968451606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,64,0.02612177696492937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,32,0.026437333888477747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,2560,0.04020177655749851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,256,0.01827111177974277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,768,0.02190133266978794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,128,0.01791733337773217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,512,0.034656001461876765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,65536,0.7587582270304362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,65536,1.9444043901231554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,51200,1.4621529049343531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,51200,0.5936915609571669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,16384,0.5088204277886285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,16384,0.5364213519626194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,16384,0.16932977570427787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,16384,0.505783134036594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,65536,3.0534364912245007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,12288,0.13637777169545492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,12288,0.3328008916642931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,10240,0.3260426786210802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,51200,2.1522017584906683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,65536,2.074250751071506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,51200,1.6141013039482965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,8192,0.24009866184658477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,10240,0.29536888334486217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,12288,0.3799911075168186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,8192,0.08836355474260117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,8192,0.21967644161648223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,7168,0.07821333408355713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,7168,0.19509511523776582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,6144,0.17477333545684814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,12288,0.34536178906758624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,6144,0.06843822532229953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,6144,0.18882221645779082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,8192,0.24791288375854492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,6144,0.16893511348300508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,5120,0.059676441881391734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,5120,0.15856622325049505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,4096,0.11737422148386638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,5120,0.14358399973975286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,10240,0.32088621457417804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,4096,0.05028444528579712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,4096,0.1269351111518012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,4096,0.11912266413370769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,7168,0.22122489081488717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,3584,0.04522577921549479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,3584,0.11215821901957194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,3584,0.10612444745169745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,7168,0.21789688534206816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,3072,0.039870222409566246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,3072,0.0987173318862915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,3072,0.0936631096733941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,3584,0.09947644339667426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,2560,0.08443466822306316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,2048,0.06047111087375217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,2560,0.08132355742984347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,5120,0.1396791140238444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,1536,0.04453688859939575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,2048,0.0302471121152242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,2048,0.070214225186242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,2048,0.0699911117553711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,1024,0.03245155678855048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,1536,0.05573155482610067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,1536,0.056399113602108426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,10240,0.11085955301920573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,768,0.025801777839660645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,1024,0.04253422220547994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,1024,0.021827555365032617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,2560,0.035375111632876925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,1024,0.04348977737956577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,768,0.01941422290272183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,768,0.03449777762095133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,3072,0.09190311034520467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,512,0.028212444649802312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,1536,0.025248888466093276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,2560,0.07721510860655043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,256,0.02477600011560652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,128,0.01202044470442666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,256,0.016418667303191293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,256,0.023952000670962866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,512,0.018784888916545443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,64,0.011090666883521609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,128,0.023338667220539514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,32,0.011402666568756104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,768,0.03718488746219211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,256,0.014292443792025248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,64,0.02316711180739933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,512,0.030730668041441176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,32,0.023046221998002794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,128,0.016294222739007738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,65536,0.6365626653035482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,512,0.017490666773584154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,51200,1.2125039630466037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,65536,1.883768081665039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,16384,0.4118693404727512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,51200,0.5117395718892416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,128,0.020503999458418954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,16384,0.1666062275568644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,16384,0.4654826588100857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,12288,0.2657866742875841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,16384,0.409060451719496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,51200,2.0417537689208984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,51200,1.3605288399590387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,10240,0.2524248758951823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,12288,0.2951084507836236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,10240,0.10827110873328315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,8192,0.20649777518378365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,10240,0.2552533414628771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,65536,2.5926905737982855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,8192,0.07664177815119426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,8192,0.220176895459493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,8192,0.2054506672753228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,7168,0.19053955872853598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,7168,0.0690844456354777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,7168,0.18884710470835367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,7168,0.17704621950785318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,65536,1.8227645026312933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,6144,0.1625208854675293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,6144,0.061533331871032715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,6144,0.15428000026279026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,12288,0.33475645383199054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,5120,0.05272622240914238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,5120,0.1352862252129449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,5120,0.13187732961442736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,10240,0.2956648932562934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,4096,0.043635557095209755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,4096,0.10889244741863674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,3584,0.09300266371832953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,4096,0.10934311151504517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,12288,0.11135200659434001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,3584,0.039672000540627375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,3584,0.09656000137329102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,3072,0.07413778040144178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,3584,0.09613955683178371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,2560,0.06443822383880615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,3072,0.03432888785998026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,4096,0.10168266958660549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,6144,0.16356533103519016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,2048,0.05167822374237908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,2560,0.03068000078201294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,2560,0.07288089063432482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,2560,0.07346044646369086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,2048,0.02628444466325972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,5120,0.11643821663326687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,2048,0.06146933635075887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,1024,0.029839111698998347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,1536,0.04864711231655545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,3072,0.08545600043402778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,3072,0.08475289079878066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,1024,0.03661511010593838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,1024,0.01924800044960446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,1536,0.039600001441107854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,2048,0.06042755312389798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,512,0.017316444052590262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,768,0.017330666383107502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,768,0.03321866525544061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,512,0.024510223004553053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,1536,0.05047288868162367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,256,0.01387733303838306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,768,0.02292355563905504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,256,0.02124711043304867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,256,0.021664889322386846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,256,0.01516266663869222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,768,0.029735111527972754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,128,0.02027555637889438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,128,0.014892444014549255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,64,0.012822222378518848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,1536,0.022656889425383672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,512,0.027495111028353374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,64,0.020019556085268658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,32,0.019985778464211356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,1024,0.03969066672854953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,128,0.011767999993430244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,128,0.018230222993426852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,32,0.013015111287434896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,65536,0.5282613436381022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,65536,1.3683172861735027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,51200,1.0889964633517795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,16384,0.3561413288116455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,51200,0.4260479874081082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,65536,1.86314943101671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,16384,0.1286862161424425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,65536,1.4033181932237413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,512,0.016178665889634024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,12288,0.27530932426452637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,51200,1.1114258236355252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,12288,0.09435555669996475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,10240,0.22133155663808188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,12288,0.2716097831726074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,12288,0.24785510698954263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,8192,0.1557919979095459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,10240,0.2751644452412923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,10240,0.20184177822536895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,8192,0.17780000633663604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,7168,0.1586079994837443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,8192,0.16411556137932673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,16384,0.33025066057840985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,51200,1.3871929380628798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,6144,0.13188355498843723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,7168,0.05806666612625122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,7168,0.14641866419050428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,6144,0.05130755570199755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,10240,0.08182489209704928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,6144,0.13434667057461208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,6144,0.12593866719139948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,5120,0.043839110268486865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,16384,0.39073954688178164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,5120,0.10879377524058025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,4096,0.09137511253356934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,7168,0.1566471126344469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,4096,0.08939288722144233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,3584,0.07857155799865723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,5120,0.11210132969750299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,3584,0.08085599872801039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,4096,0.08750044637256199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,3072,0.07161333163579305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,3072,0.07088177733951144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,3072,0.07074844174914889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,3072,0.02959644463327196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,2560,0.05641688903172811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,3584,0.03316088848643833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,3584,0.08050755659739177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,2560,0.026388444834285315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,4096,0.036489778094821505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,2048,0.05104000038570828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,1536,0.0353999998834398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,1536,0.04085777865515815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,8192,0.06509600083033244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,1536,0.019710222880045574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,1536,0.04275022281540764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,2048,0.022813333405388728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,1024,0.03107288810941908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,1024,0.016745777593718637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,1024,0.03346755438380771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,768,0.01834133267402649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,2560,0.061230222384134926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,768,0.015485333071814643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,768,0.025303110480308533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,512,0.016178665889634024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,768,0.027775110469924078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,2048,0.05194488830036587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,512,0.014479110638300577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,512,0.020863110820452373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,512,0.023584889041052923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,2560,0.06115111377504137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,256,0.01791733337773217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,256,0.013597333596812354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,5120,0.11338222026824951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,128,0.013502221968438892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,1024,0.025800888737042744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,128,0.01724177764521705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,32,0.008519111408127679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,256,0.01276000009642707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,32,0.01717422240310245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,128,0.00942133367061615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,2048,0.04457066787613762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,128,0.017253332667880587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,65536,1.1943564944797092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,51200,0.7122124565972222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,256,0.0192586663696501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,65536,1.642864015367296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,65536,1.2883520126342773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,51200,0.3681306574079726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,65536,0.4512106577555339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,16384,0.10781688822640313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,51200,1.525970670912001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,64,0.01740711099571652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,16384,0.29564976692199707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,64,0.008396444221337637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,12288,0.23943734169006348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,10240,0.1538355615403917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,16384,0.2910044458177355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,10240,0.07309689124425252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,10240,0.20324444770812988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,8192,0.157242668999566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,12288,0.21829422314961752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,8192,0.0579013360871209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,12288,0.08540178007549709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,7168,0.12378133667839898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,8192,0.1456728908750746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,12288,0.21380355623033312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,6144,0.11016888750924005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,16384,0.3188400003645155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,7168,0.13268178039126927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,10240,0.18053244219885933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,6144,0.12202221817440456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,5120,0.07961955335405138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,6144,0.11334844430287679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,5120,0.03874666823281182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,5120,0.10582133134206136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,4096,0.07020266850789388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,7168,0.05237422386805216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,4096,0.08374044630262588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,4096,0.03203999996185303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,3584,0.06495910882949829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,6144,0.04577689038382637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,8192,0.15996710459391275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,3584,0.029836446046829224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,3584,0.07254133621851604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,5120,0.0983004437552558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,7168,0.14111733436584473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,2560,0.04383555716938443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,3072,0.02661688956949446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,4096,0.08043911059697469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,2560,0.0543120006720225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,2048,0.034267554680506386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,3584,0.07406222158008151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,2048,0.046126223272747464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,3072,0.06421955426534016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,2048,0.04688622223006355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,3072,0.06407022476196289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,1536,0.026657778355810378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,51200,0.9700764550103081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,2560,0.056185775332980685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,1024,0.020466667082574632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,1536,0.03695288962788052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,1536,0.03847288754251268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,3072,0.05541066659821404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,768,0.016207999653286405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,1024,0.01586844523747762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,1024,0.030403554439544678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,768,0.022843556271659002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,2048,0.02085777786042955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,512,0.012304888831244575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,512,0.018992889258596633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,512,0.021634666456116572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,256,0.009683555199040307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,512,0.01346577786737018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,1024,0.027726221415731642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,256,0.01276000009642707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,768,0.014731556177139282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,768,0.025277333127127752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,128,0.009010666774378883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,64,0.00794400026400884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,128,0.015670221712854173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,2560,0.024029332730505202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,128,0.01612622208065457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,64,0.015484443969196744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,32,0.015839111473825242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,256,0.016527111331621807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,256,0.01791288952032725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,128,0.012814222110642327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,65536,1.0190835528903537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,1536,0.01831999917825063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,65536,0.43487734264797634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,51200,0.6759022076924642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,51200,0.33184978697035045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,65536,1.279963599310981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,65536,1.0876524183485243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,16384,0.09938222169876099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,12288,0.20825510554843477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,16384,0.29181332058376735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,51200,0.996122678120931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,51200,0.8084479967753092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,16384,0.2511884371439616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,12288,0.07887022362815009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,10240,0.17816355493333605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,12288,0.20964533752865264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,12288,0.19043555524614122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,16384,0.2684231069352892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,10240,0.1745591163635254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,10240,0.15627643797132704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,8192,0.05379822187953525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,8192,0.14127555158403185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,8192,0.128711117638482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,7168,0.04759111007054647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,7168,0.12563289536370173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,32,0.008273777862389883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,7168,0.1140693293677436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,5120,0.08080799712075128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,6144,0.04127911064359877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,6144,0.10723377598656549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,8192,0.14501244492001003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,4096,0.0697751111454434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,5120,0.03503200080659654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,7168,0.1171297762129042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,5120,0.08574666579564412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,3584,0.05513599846098158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,4096,0.02947288751602173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,6144,0.10340622398588394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,3584,0.025967111190160114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,3584,0.06554133362240262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,3072,0.05240799983342489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,3584,0.06389955679575603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,10240,0.06708266337712605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,2560,0.04060622387462192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,3072,0.023589332898457844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,3072,0.05676533116234673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,5120,0.09153155485788982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,3072,0.056927998860677086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,2048,0.03522400061289469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,2560,0.021869333253966436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,2560,0.048976888259251915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,4096,0.07360000080532499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,1536,0.025957332717047796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,2560,0.04907466636763679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,2048,0.04168622361289131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,1024,0.017330666383107502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,1536,0.016975111431545682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,1536,0.03360622127850851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,1536,0.034493333763546415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,6144,0.09968355629179214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,768,0.014594667487674289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,1024,0.014524444937705994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,1024,0.0266275554895401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,768,0.013627555635240344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,768,0.0207342223988639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,768,0.02269777821169959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,4096,0.07357155614429049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,512,0.017266665895779926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,256,0.0090524446633127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,512,0.01959288948112064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,256,0.015695111619101632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,1024,0.025033776958783466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,128,0.007660444411966536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,256,0.016890666551060148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,256,0.01238666640387641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,512,0.011788444386588203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,64,0.007333333293596904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,128,0.011923555698659686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,128,0.014145778285132514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,2048,0.018968888454967074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,512,0.012793777717484368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,64,0.013935999737845527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,2048,0.04195555713441637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,32,0.01447111037042406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,65536,0.3728480074140761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,65536,0.879491541120741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,51200,0.6514524353875054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,32,0.007648000286685095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,128,0.01517955627706316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,16384,0.22266578674316406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,51200,0.3554684321085612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,65536,1.0113111072116427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,16384,0.11785599920484756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,16384,0.24032176865471733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,16384,0.24491911464267305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,51200,0.7864178021748861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,12288,0.07028444608052571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,12288,0.1906942261589898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,12288,0.19031200144026014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,10240,0.0621395574675666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,8192,0.10950399769677056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,10240,0.16159021854400635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,65536,1.0976613362630208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,10240,0.15630133946736655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,7168,0.09270755449930827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,8192,0.04902044600910611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,8192,0.12367022037506104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,12288,0.20030311743418375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,7168,0.043523556656307645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,7168,0.10866844654083252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,7168,0.11135021845499675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,10240,0.13840888606177434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,5120,0.06534577740563287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,6144,0.09535555707083808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,6144,0.09778489006890191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,4096,0.05137066708670723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,5120,0.08071200052897136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,5120,0.08262844218148126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,4096,0.027064889669418335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,8192,0.12689867284562853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,6144,0.08208444383409289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,4096,0.06955999798244901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,6144,0.03746933407253689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,3072,0.037147555086347796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,5120,0.03235200047492981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,3072,0.05098311106363932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,51200,0.8364195293850369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,3584,0.0470728874206543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,2560,0.03277777632077535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,3584,0.024890666206677754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,2560,0.043431109852261014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,3584,0.06203822294871012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,2048,0.025868443979157343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,2560,0.04871822065777249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,2048,0.018128888474570382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,2048,0.03732444511519538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,2048,0.04020355476273431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,4096,0.06463911135991414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,3072,0.054935110939873584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,1024,0.015235554840829639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,1536,0.029826667573716905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,1536,0.015658666690190632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,1536,0.03315466642379761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,768,0.012608889076444836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,1024,0.022247110803922016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,2560,0.020260444945759244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,1024,0.025991999440722998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,512,0.010120000276300643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,768,0.013281777501106262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,1536,0.021873777111371357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,3072,0.022851554883850947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,512,0.01257333325015174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,512,0.018719111879666645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,3584,0.05743733379575941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,1024,0.013747555514176687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,128,0.007541333635648091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,256,0.011745777395036487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,768,0.01868355605337355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,128,0.01276711126168569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,128,0.0138568886452251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,512,0.015604444675975375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,256,0.008029333419269985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,64,0.012466666599114736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,256,0.01347555551264021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,256,0.015633778439627755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,65536,0.6768515374925402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,64,0.00655377780397733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,32,0.006659555352396435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,65536,0.3213137785593669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,32,0.01276088919904497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,51200,0.5189769003126357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,768,0.02165866725974613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,128,0.011425777441925473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,16384,0.17351999547746447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,65536,0.8858622445000542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,16384,0.2068586614396837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,16384,0.07858933342827691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,51200,0.25006045235527885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,12288,0.15484000576867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,51200,0.6936995718214246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,10240,0.10743466350767349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,12288,0.0614853302637736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,51200,0.6432799763149685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,10240,0.053448001543680825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,8192,0.09036799934175278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,16384,0.2087519963582357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,10240,0.1343982219696045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,7168,0.07827644215689765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,8192,0.04241066508822971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,8192,0.1056542197863261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,8192,0.10956355598237778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,6144,0.06755110952589247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,7168,0.09284444650014241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,7168,0.09655644496281941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,12288,0.15782399972279867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,6144,0.08083555433485243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,10240,0.13156267007191977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,6144,0.08613866567611694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,5120,0.028796444336573284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,5120,0.06821155548095703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,4096,0.0439591109752655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,5120,0.07218133078681098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,7168,0.03675999906327989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,4096,0.023929778072569106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,3584,0.03840355409516229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,4096,0.054914666546715625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,4096,0.06140977806515164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,3072,0.0344888899061415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,5120,0.05597866906060112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,3584,0.04882488979233635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,65536,0.8894969092475044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,2560,0.02958844436539544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,3072,0.04355822337998284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,3072,0.048024889495637685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,2560,0.03746933407253689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,2048,0.024311999479929607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,2560,0.018629334039158292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,3584,0.022112000319692824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,6144,0.03237955437766181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,1536,0.01940000057220459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,3584,0.05389155613051521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,3072,0.020096888144810993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,1536,0.014818666709793938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,1024,0.015174221661355762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,1536,0.029418667157491047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,2560,0.04270222120814853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,1024,0.023048889305856492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,768,0.0120000003112687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,2048,0.03159644537501865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,2048,0.01663466625743442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,2048,0.03549777799182468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,512,0.009460444251696268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,768,0.016466667254765827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,1536,0.025790222816997107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,768,0.01996799972322252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,512,0.011408000356621213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,512,0.013737777868906656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,512,0.016884444488419425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,1024,0.019466666711701285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,128,0.006923555499977536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,256,0.010932444698280759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,1024,0.013119999733236102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,12288,0.13666044341193304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,64,0.007288888924651676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,128,0.010845333337783813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,128,0.011050666371981302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,768,0.012123555772834353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,128,0.012644444074895648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,64,0.010743111371994019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,32,0.011064888702498542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,65536,0.4335333241356744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,256,0.011745777395036487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,65536,0.2551822132534451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,256,0.013560000393125745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,65536,0.6876382297939725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,16384,0.13495288954840765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,65536,0.6659590933057997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,51200,0.5495484670003256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,256,0.007236444287829929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,51200,0.5174773534138998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,12288,0.09436711337831284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,16384,0.06843466891182794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,16384,0.17063021659851074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,10240,0.0892595582538181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,12288,0.05595644315083822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,12288,0.12963910897572836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,12288,0.13077244493696424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,32,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,10240,0.04797600044144524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,10240,0.10940355724758571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,8192,0.0634933312733968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,10240,0.1106577780511644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,7168,0.05668533510631985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,8192,0.03520355621973673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,16384,0.17184621757931182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,8192,0.09047200282414754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,6144,0.04864177770084805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,7168,0.07799910836749606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,51200,0.34836888313293457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,7168,0.08228266901440091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,5120,0.04006310966279771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,6144,0.028548445966508653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,51200,0.20681688520643446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,6144,0.07090310917960273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,4096,0.030497776137457952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,5120,0.05652177996105618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,5120,0.024544888072543677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,5120,0.06077600187725491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,4096,0.02055377761522929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,3584,0.027319999204741582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,7168,0.03154933452606201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,3584,0.018980445133315194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,3584,0.04561688833766513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,6144,0.06709688901901245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,3072,0.036900444163216486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,2560,0.020632889535692003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,3072,0.017494221528371174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,4096,0.04626489016744825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,4096,0.05189333359400431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,2048,0.017210667332013447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,2560,0.03201866812176175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,3584,0.04140355520778232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,2560,0.03580800029966567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,3072,0.022861333356963262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,1536,0.013719999955760108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,2048,0.014901333385043673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,2048,0.026714666022194758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,1024,0.01071911139620675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,1536,0.01316088851955202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,1536,0.022034666604465906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,3072,0.04044355617629157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,8192,0.08801511261198257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,768,0.008678221868144141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,1024,0.011752000285519494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,1024,0.020036444067955017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,768,0.011125333607196808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,2560,0.01592088904645708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,768,0.017218665944205392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,512,0.012156444291273752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,256,0.005963555640644497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,2048,0.029947555727428857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,1536,0.02493422230084737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,256,0.010336000058386061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,128,0.005274666680230035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,512,0.006996444529957241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,768,0.014130666851997375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,128,0.009689778089523315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,128,0.010040000081062317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,64,0.004950222041871813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,512,0.010399999717871347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,32,0.005092444519201915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,512,0.014839111102951897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,64,0.009368889033794403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,32,0.009355555805895064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,65536,0.3553982310824924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,1024,0.016647110382715862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,256,0.010037333601050908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,51200,0.2791217697991265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,65536,0.5490773518880209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,16384,0.0907813310623169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,65536,0.5164755715264214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,51200,0.4201360013749864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,128,0.011735111474990845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,12288,0.07154044177797106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,16384,0.05592711104287041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,16384,0.13557067182328966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,256,0.012480888929631976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,12288,0.10534755388895671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,10240,0.05699377589755588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,12288,0.04389155573315091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,12288,0.10535644160376655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,8192,0.04620088802443611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,10240,0.038415110773510404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,10240,0.0884595579571194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,10240,0.08886666430367364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,7168,0.04009955459170871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,8192,0.028274665276209515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,16384,0.13963822523752847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,8192,0.0732791092660692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,6144,0.03374044431580438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,7168,0.02531288895342085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,7168,0.06215111414591471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,7168,0.06463911135991414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,65536,0.2057173384560479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,5120,0.028911110427644517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,6144,0.05416622095637851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,6144,0.05708977911207411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,5120,0.019672888848516677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,5120,0.04567111200756497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,4096,0.02385422256257799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,5120,0.0490551127327813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,51200,0.4034630987379286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,3584,0.02206755512290531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,4096,0.017248888810475666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,4096,0.037829332881503634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,4096,0.04088266690572103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,3072,0.0195795562532213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,3584,0.037019554111692644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,6144,0.022595556245909795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,3072,0.030135111676322088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,2560,0.01645955608950721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,8192,0.07093333535724215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,2560,0.02608355548646715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,2560,0.029030223687489826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,3584,0.034197333786222674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,2048,0.014239999983045789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,2048,0.022328888376553852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,3072,0.0151591118839052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,2048,0.024480000138282776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,1536,0.011976000335481433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,3072,0.03318399853176541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,2048,0.012434666355450949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,1024,0.009736888938479954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,1536,0.011411555939250521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,1536,0.018410666121376883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,768,0.008555555509196388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,1536,0.020608888732062448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,2560,0.013793778088357715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,1024,0.013983110586802164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,512,0.00700800038046307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,768,0.010086221827401055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,768,0.012096888489193387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,768,0.014543111125628153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,256,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,512,0.013312888642152151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,256,0.00869511150651508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,3584,0.015844444433848064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,256,0.011408000356621213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,128,0.005234666582610872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,1024,0.016551999582184684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,128,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,64,0.004573333180612988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,64,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,32,0.004596444467703502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,1024,0.01071999967098236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,256,0.009191110730171204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,32,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,65536,0.283735990524292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,128,0.009024000002278222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,128,0.01073333372672399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,51200,0.20274844434526232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,65536,0.45358578364054364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,51200,0.15837777985466853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,51200,0.36224712265862363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,16384,0.06807555754979451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,51200,0.4080933464898004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,51200,0.17655911710527208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,16384,0.11819911003112793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,16384,0.05485599901941088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,16384,0.13478399647606745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,65536,0.21408266491360137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,10240,0.04676977793375651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,12288,0.0433555543422699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,12288,0.10263199938668145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,65536,0.5112977557712132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,8192,0.034824000464545354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,10240,0.0758426653014289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,10240,0.03773599863052368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,10240,0.08778133657243516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,7168,0.0284897784392039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,8192,0.06064089139302572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,512,0.009666666388511658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,8192,0.07211377885606554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,6144,0.02511288887924618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,7168,0.025218667255507574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,7168,0.053583999474843345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,12288,0.08938666847017075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,7168,0.06366666820314196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,6144,0.022294221652878657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,6144,0.046784000264273755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,6144,0.05671911107169258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,5120,0.01961599952644772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,4096,0.017245334055688646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,8192,0.027549333042568628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,5120,0.048248890373441905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,512,0.010374222364690568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,4096,0.01714666684468587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,3584,0.02942488922013177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,3584,0.015951111912727356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,3584,0.03567288981543647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,5120,0.039804445372687444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,3072,0.025901334153281316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,3072,0.03181688984235128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,3072,0.014801777071423002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,4096,0.03277511066860623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,5120,0.019870221614837646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,2560,0.02271644439962175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,2560,0.013738666971524557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,2560,0.02826666831970215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,2048,0.010378667049937779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,2048,0.019146665930747986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,2048,0.02387999991575877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,1536,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,2048,0.01255644443962309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,4096,0.04008177916208903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,1536,0.011410666836632622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,1536,0.01611555616060893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,1536,0.020543111695183646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,3072,0.013493333425786761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,1024,0.012430222498046027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,1024,0.010545777777830759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,1024,0.016515556308958266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,2560,0.012039110892348819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,768,0.010038221875826517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,512,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,768,0.014482667048772177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,1024,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,256,0.004587555511130227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,512,0.009696000152164036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,512,0.009372444616423713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,512,0.013280889226330651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,128,0.003929777691761653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,768,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,12288,0.05085066623157925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,128,0.007322666545708974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,64,0.003875555677546395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,64,0.007191110816266801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,32,0.0038773334688610504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,256,0.00906133320596483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,256,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,3584,0.015599111715952555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,32,0.007320889168315464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,768,0.010710222025712332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,128,0.01036533299419615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,65536,0.15345866150326198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,51200,0.1527493397394816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,65536,0.4009973208109538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,16384,0.05078044533729553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,65536,0.3616666793823242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,51200,0.31602133644951713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,51200,0.28498045603434247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,16384,0.10534755388895671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,256,0.011425777441925473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,65536,0.19478044244978163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,16384,0.09772800074683295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,10240,0.03235288792186313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,12288,0.03460977805985345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,12288,0.08018222120073107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,12288,0.07507555352316962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,8192,0.026007999976476032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,10240,0.029863112502627905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,51200,0.12280000580681695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,10240,0.0639680027961731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,7168,0.021968000464969214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,8192,0.021336000826623704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,8192,0.054486221737331815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,12288,0.039164443810780845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,16384,0.04293333490689596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,6144,0.01867644488811493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,7168,0.01904800037542979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,7168,0.04805244339836968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,128,0.00908622228437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,5120,0.016947555873129103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,7168,0.047001779079437256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,6144,0.041991998751958214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,4096,0.01405066748460134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,5120,0.015274667077594332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,5120,0.036004443963368736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,5120,0.035938666926489934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,3584,0.012464889221721225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,4096,0.01350400017367469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,4096,0.02953333324856228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,4096,0.029602666695912678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,3072,0.011446221835083432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,3584,0.012784000072214337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,3584,0.026320000489552815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,3584,0.027097778187857732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,2560,0.010048888623714447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,3072,0.012084444363911947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,3072,0.023756444454193115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,3072,0.024503111839294434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,6144,0.04157155421045091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,2560,0.011407111254003314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,10240,0.06726133161120944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,2560,0.021680000755521987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,1536,0.007671111159854465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,2048,0.017786666750907898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,2048,0.01889511115021176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,8192,0.0537377761469947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,1024,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,1536,0.01459111107720269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,1536,0.01590666671593984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,6144,0.01706755492422316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,768,0.005528888768619961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,1024,0.011417778001891242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,2048,0.0086666668454806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,1024,0.013262222210566202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,2560,0.020765332712067496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,768,0.009032888544930352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,768,0.010026666853162978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,2048,0.010743111371994019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,768,0.011815999945004782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,256,0.004218666503826777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,512,0.00867733359336853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,1536,0.010099555883142684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,256,0.007258666886223688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,256,0.00831200016869439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,1024,0.009480000370078618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,128,0.006747555401590135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,128,0.009006222089131674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,128,0.008299555215570662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,512,0.00868444475862715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,32,0.004071111066473855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,512,0.010728889041476779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,64,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,65536,0.11527110470665826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,128,0.0042053332759274375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,256,0.009455111291673448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,51200,0.09387999773025513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,65536,0.15182222260369196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,65536,0.3336702187856038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,16384,0.03839200072818332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,65536,0.34755200809902614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,51200,0.12156800429026286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,51200,0.2550506591796875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,12288,0.028228445185555354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,16384,0.042592889732784696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,16384,0.08754222260581122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,16384,0.09300888909233941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,32,0.006640000061856375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,12288,0.06483199861314562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,12288,0.07152977916929457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,512,0.004942222187916438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,10240,0.02955289019478692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,10240,0.055388443999820285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,10240,0.061352001296149365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,51200,0.2724871105617947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,8192,0.02064622276359134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,7168,0.015513777732849121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,8192,0.050833778248892895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,12288,0.03394133183691237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,6144,0.013812444276279874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,7168,0.045034666856129967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,8192,0.017117333081033494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,6144,0.03444177905718485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,64,0.004252444538805219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,6144,0.01700266698996226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,8192,0.04418666826354133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,5120,0.029483556747436523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,4096,0.010070222119490305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,10240,0.02052977846728431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,7168,0.039311110973358154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,4096,0.013510222236315409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,3584,0.009488888912730748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,4096,0.02834577692879571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,5120,0.012246222131782107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,6144,0.03940800163480971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,3072,0.00851733320289188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,3584,0.01279022213485506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,3584,0.022044445077578228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,3584,0.025787555509143408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,5120,0.015156444576051501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,2560,0.007669332954618666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,3072,0.012116444607575735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,7168,0.018533334136009216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,2048,0.0069520001610120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,2560,0.011462222370836469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,2560,0.01722133325205909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,4096,0.024508444799317256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,1536,0.006224000205596288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,2048,0.01092177795039283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,2048,0.014857777290874057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,2048,0.01793066660563151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,1536,0.010346666806273991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,1536,0.012447111308574677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,1536,0.015185778339703878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,1024,0.009354666703277165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,768,0.004564444637960858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,2560,0.020851555797788832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,768,0.008911111288600499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,768,0.009045333498054082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,1024,0.004936000125275718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,512,0.004212444441186057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,1024,0.010033778018421596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,3072,0.023411555422676936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,256,0.0035786665976047516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,256,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,768,0.011314666933483548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,256,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,256,0.009364444348547194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,128,0.003524444583389494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,512,0.010729778144094678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,512,0.00868444475862715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,64,0.003245333416594399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,128,0.005684444473849402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,32,0.0032115555885765287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,128,0.008529778156015608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,64,0.005628444254398346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,32,0.005592888842026393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,65536,0.08925511439641316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,5120,0.034373333056767784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,1024,0.012744888663291931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,65536,0.29131555557250977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,51200,0.07224622037675646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,65536,0.15144444836510554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,65536,0.3427715566423204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,16384,0.029106666644414265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,51200,0.12110666433970134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,51200,0.22847466998630098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,51200,0.2688346703847249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,12288,0.02065955599149068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,3072,0.019727110862731934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,128,0.008086221913496653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,12288,0.07049066490597196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,10240,0.01645777788427141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,16384,0.07645866605970594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,16384,0.04283377859327528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,10240,0.04803911182615492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,10240,0.029480000336964924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,10240,0.06025155385335287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,12288,0.05691822369893392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,8192,0.021648888786633808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,8192,0.038945777548684016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,8192,0.049807998869154185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,7168,0.019879110985332064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,7168,0.034633778863483004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,6144,0.010612444745169746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,7168,0.04404711061053806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,512,0.007663999994595845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,6144,0.030014223522610132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,5120,0.009407999614874521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,6144,0.03865955604447259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,7168,0.011953777737087674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,5120,0.02587288949224684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,5120,0.0333262218369378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,12288,0.034811556339263916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,4096,0.02164977788925171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,3584,0.007401777638329401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,4096,0.013982221484184265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,4096,0.0282008879714542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,8192,0.013384000294738345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,3072,0.006959111326270633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,3584,0.013153778182135688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,3584,0.019711999429596793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,3584,0.025382222400771245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,5120,0.01591377788119846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,2560,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,3072,0.017614222235149808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,3072,0.02330577704641554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,2560,0.011765333513418833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,2560,0.01550222271018558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,2560,0.02013777858681149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,2048,0.011100444528791638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,1536,0.00489066665371259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,2048,0.017627555463049147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,1536,0.01144444445768992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,1536,0.014861333701345654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,1024,0.004238222208287981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,1536,0.01038755559258991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,1024,0.009249777429633671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,6144,0.01789777808719211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,768,0.0038817777401871155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,3072,0.012456888953844706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,2048,0.005323555734422471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,768,0.00908799966176351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,512,0.0035377778112888336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,16384,0.0920577777756585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,1024,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,512,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,256,0.0031893334041039148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,4096,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,256,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,256,0.009019555317031013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,128,0.0028728888266616394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,2048,0.0135057775510682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,128,0.0052879999081293745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,512,0.008676444490750631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,64,0.0028888889484935333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,128,0.008676444490750631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,768,0.00816888858874639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,32,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,65536,0.08665777577294244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,64,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,32,0.005316444569163852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,51200,0.06695289081997342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,16384,0.026775999201668635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,1024,0.012436444560686747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,12288,0.016567111015319824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,65536,0.2911777761247423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,10240,0.01575911045074463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,51200,0.22251645723978677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,16384,0.07629155450397067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,8192,0.012797333300113678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,512,0.010378667049937779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,128,0.008295111358165741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,6144,0.01038044442733129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,10240,0.048771556880739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,8192,0.03923644291030036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,5120,0.009004444711738164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,768,0.011412444214026133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,4096,0.007692444655630324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,5120,0.0255084451701906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,3584,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,4096,0.02168888847033183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,3072,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,7168,0.011415111521879831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,2560,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,3584,0.019504888190163504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,2048,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,3072,0.01763199932045407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,1536,0.004577777865860197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,2560,0.01552177800072564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,256,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,2048,0.013473778135246701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,768,0.003900444342030419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,1536,0.011409777734014722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,512,0.0035297779573334586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,1024,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,256,0.0032355555643637977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,768,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,128,0.002954666606254048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,512,0.007295110987292395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,64,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,256,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,7168,0.03432177835040622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,128,0.005294222384691238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,6144,0.029987557066811457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,1024,0.0038817777401871155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,64,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,51200,0.0689386659198337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,32,0.0052897776994440295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,16384,0.02538577715555827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,12288,0.01315111087428199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,10240,0.011781333221329583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,65536,0.2849751048617893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,12288,0.05676266882154676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,16384,0.07606755362616645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,12288,0.05669422282112969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,10240,0.04797600044144524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,32,0.002893333426780171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,7168,0.03433777888615926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,65536,0.08367644415961371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,6144,0.030208001534144085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,4096,0.00980444418059455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,5120,0.025407999753952026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,3584,0.009020444419648912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,8192,0.010384889112578498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,7168,0.009684444301658207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,4096,0.021756443712446425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,2560,0.0069875551594628235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,3584,0.019832000136375427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,2048,0.0053048887186580235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,3072,0.017895999881956313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,5120,0.011367110742463006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,2560,0.01588800052801768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,1024,0.0042160000238153665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,2048,0.01348533315791024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,768,0.003551111039188173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,1536,0.011397333608733283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,512,0.003536000019974179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,1024,0.009399111072222391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,3072,0.007976888782448238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,768,0.00814666681819492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,128,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,512,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,64,0.0029155556112527847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,256,0.005783111270931032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,32,0.0029120000286234748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,128,0.0052728888889153796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,64,0.005244444641802046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,32,0.004965333475006951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,51200,0.22219999631245932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,256,0.0032071111102898917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,6144,0.008943999807039896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,1536,0.004563555535342959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,8192,0.038782222403420344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,16384,1.5739235348171656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,12288,1.1401910781860352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,12288,3.5483769310845266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,16384,5.1082844204372835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,12288,3.661054399278429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,16384,4.419485304090712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,10240,0.9647057851155599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,51200,6.283641815185547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,10240,3.0135031806098094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,10240,3.015714645385742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,8192,2.218409856160482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,8192,0.7330711152818469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,8192,2.2661209106445312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,7168,0.6878462367587619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,6144,0.5739084349738227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,7168,1.9867911868625219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,6144,1.8594871097140844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,6144,1.7241235309176977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,5120,0.5009173287285699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,5120,1.6128070619371202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,5120,1.4200098249647353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,4096,1.1776373121473525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,4096,0.46054665247599286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,4096,1.09733952416314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,3584,0.4060426553090413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,3584,1.2135848999023438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,7168,2.3138995700412326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,3584,0.9850586785210503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,3072,0.34733245107862687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,3072,0.9086347156100802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,2560,0.2849617799123128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,2560,0.8105555640326606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,2560,0.7094560199313693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,51200,14.255137125651041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,2048,0.582807117038303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,3072,0.8271999888949924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,2048,0.28691911697387695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,1536,0.43082311418321395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,1024,0.34544889132181805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,1024,0.177894221411811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,768,0.23623556560940215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,768,0.13341066572401258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,768,0.2833137777116564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,1536,0.4644097752041287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,256,0.12172178427378337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,512,0.11938844786749946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,512,0.2333279980553521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,1536,0.18839289082421196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,2048,0.6009635395473903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,256,0.10416889190673828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,256,0.2108915514416165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,128,0.10217866632673477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,128,0.19827732774946424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,32,0.08707200156317817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,512,0.1721724404229058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,64,0.20389066802130806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,32,0.20486577351888022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,128,0.09961689180798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,64,0.08392266432444255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,1024,0.3425315486060248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,16384,3.427009794447157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,51200,22.044420030381943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,51200,4.466310289171007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,16384,1.213754653930664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,65536,6.251041836208767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,12288,0.8957911597357856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,16384,3.97837405734592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,12288,3.064475589328342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,12288,2.949098587036133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,10240,0.7253751224941678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,51200,10.896852281358507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,10240,2.3052026960584855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,10240,2.893932342529297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,8192,1.950391133626302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,8192,0.5829528702629937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,8192,1.8294213612874348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,7168,0.5094568994310167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,7168,1.5758328967624242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,65536,13.995516459147135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,6144,0.44571553336249453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,6144,1.6207617653740778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,7168,1.7458409203423395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,5120,0.3966355588701036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,5120,1.22980711195204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,4096,0.31516533427768284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,4096,0.8607609007093641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,5120,1.0925484763251412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,6144,1.3381101820203993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,3584,0.888687981499566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,3584,0.2868533399369982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,3584,0.7774613168504504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,3072,0.2556346787346734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,2560,0.5191084543863932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,3072,0.6556764708624946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,2560,0.2183368868298001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,2048,0.4376853307088216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,2048,0.1840746667650011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,51200,16.92905002170139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,2048,0.45544889238145614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,1536,0.32975467046101886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,1536,0.15174489551120335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,1024,0.25574665599399143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,1536,0.36332355605231387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,1024,0.11853067080179851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,768,0.21348355876074898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,1024,0.27812443839179146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,768,0.10587022039625381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,768,0.214592006471422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,512,0.09483999676174587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,4096,0.9876559575398763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,512,0.18390222390492758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,128,0.07894844479031034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,512,0.13999822404649523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,2560,0.557236459520128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,256,0.0910231139924791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,3072,0.7328586578369141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,64,0.07422488927841187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,32,0.07819644610087077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,256,0.08389333221647476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,32,0.16167467170291477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,128,0.15639377964867487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,128,0.0819644464386834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,256,0.16604799694485134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,65536,22.96004570855035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,64,0.16013156043158636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,65536,1.9057129753960504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,51200,1.5598284403483074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,16384,1.2466568417019315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,16384,1.339671982659234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,65536,4.686789194742839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,51200,3.5122400919596353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,16384,0.49953153398301864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,12288,0.8162453439500598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,12288,0.2877226670583089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,10240,0.2600897683037652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,8192,0.5480622185601128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,8192,0.2019368939929538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,10240,0.7580897543165418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,8192,0.6042764451768663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,7168,0.5322142177157932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,7168,0.1807884507709079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,6144,0.4726115332709418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,7168,0.5184871355692545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,12288,0.9528222613864474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,10240,0.8655822012159559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,6144,0.43610488043891055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,65536,7.3835949367947045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,5120,0.3672160042656793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,4096,0.11208889219495986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,3584,0.2303555541568332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,4096,0.28874754905700684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,3584,0.10035644637213813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,3072,0.1945279969109429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,51200,5.350791931152344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,3584,0.2614026599460178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,5120,0.13241777155134413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,2560,0.17844888899061415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,4096,0.2948977682325575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,3072,0.22246132956610787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,2048,0.14685332775115967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,2560,0.18780177169375947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,2048,0.0660604437192281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,2048,0.15488799413045248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,1536,0.12354577912224664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,1024,0.07832711272769503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,5120,0.46518399980333114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,1024,0.04444800151718987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,768,0.06823644373151991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,1024,0.0938168896569146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,2560,0.07623022132449679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,512,0.04412710997793409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,1536,0.11644444200727676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,768,0.07469599776797824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,256,0.033951110310024686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,512,0.03782044516669379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,512,0.06412444512049358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,128,0.03103999959097968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,6144,0.15818044874403211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,1536,0.053948445452584155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,256,0.05737244420581394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,128,0.03313244382540385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,32,0.025399110383457605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,128,0.054346667395697705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,64,0.055323554409874805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,32,0.05680711401833428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,768,0.0409840014245775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,256,0.03375733229849074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,64,0.027310222387313843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,3072,0.08615110980139838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,65536,1.359383159213596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,51200,1.0354017681545682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,16384,0.9573262532552084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,16384,0.4368577798207601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,51200,2.7240844302707248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,12288,0.7280355559455024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,12288,0.25554222530788845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,16384,0.8759564293755425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,12288,0.6844755808512369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,10240,0.19291288322872588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,8192,0.39122666252983945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,8192,0.1543040010664198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,10240,0.6293324364556206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,7168,0.3354569011264377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,8192,0.4419404400719537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,7168,0.13584089279174805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,51200,4.126946767171224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,7168,0.39804800351460773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,6144,0.12001689275105794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,10240,0.5404239760504829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,6144,0.3326675626966688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,4096,0.2132115496529473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,5120,0.28088532553778756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,4096,0.08425866895251805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,3584,0.19086399343278673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,4096,0.2206115590201484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,65536,6.104740566677517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,65536,3.4371403588189016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,3584,0.07710578044255574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,3072,0.16127288341522217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,3584,0.19391377766927084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,5120,0.27472978168063694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,3072,0.19130399492051867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,5120,0.1026195552614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,2560,0.14465155866410997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,2048,0.05123644404941135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,2048,0.11982756190829807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,1536,0.08527644475301106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,6144,0.3570080068376329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,1024,0.06374577681223552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,1536,0.09570133023791844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,3072,0.06934933529959784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,768,0.0513626668188307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,1024,0.03608266512552897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,1024,0.07223466369840834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,2048,0.10814044210645889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,512,0.03807288739416335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,768,0.03370577759212918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,256,0.027522666586769953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,512,0.030970666143629286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,512,0.050055109792285495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,128,0.02237866653336419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,256,0.027715555495686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,1536,0.042853332228130765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,64,0.019301333361201815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,2560,0.14020533031887478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,32,0.01923111081123352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,128,0.04233155647913615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,2560,0.05951911211013794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,64,0.04305866691801283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,32,0.04355466696951124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,768,0.05823733409245809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,256,0.04482488830884298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,128,0.02737422287464142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,51200,0.9327831268310547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,16384,0.7507200241088867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,16384,0.8016799820793999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,16384,0.3150533305274116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,12288,0.5864888827006022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,12288,0.1974088880750868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,10240,0.4541493521796332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,12288,0.5788239902920193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,10240,0.16688533624013266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,10240,0.48399024539523655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,51200,2.217704984876845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,51200,3.449634763929579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,8192,0.13476178381178114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,65536,2.84171634250217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,8192,0.3827786710527208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,65536,1.1829004287719727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,6144,0.3130684428744846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,65536,4.503239101833768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,6144,0.10085155566533406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,6144,0.2832186751895481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,8192,0.37745245297749835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,5120,0.08635110987557305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,4096,0.188408891359965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,5120,0.24859378072950575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,3584,0.16108889049953884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,4096,0.18727021747165254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,3584,0.06525422467125787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,3072,0.13980889320373535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,7168,0.3476613362630208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,3072,0.0587315559387207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,2560,0.1181759966744317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,3072,0.1448755529191759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,5120,0.2515724500020345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,2048,0.09646222326490615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,2560,0.12300622463226318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,2048,0.044284443060557045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,7168,0.32091466585795086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,7168,0.12007199393378364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,3584,0.16451199849446616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,1024,0.062055110931396484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,1536,0.037427554527918495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,768,0.046494222349590726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,1024,0.06233244472079807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,2560,0.05204888847139147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,512,0.03446400165557861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,768,0.029156443145540025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,768,0.049954666031731494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,256,0.023746665981080797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,2048,0.10183022419611613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,512,0.042464888758129544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,256,0.02409155501259698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,256,0.037815110550986394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,1024,0.031367109881507024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,64,0.01651022169325087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,4096,0.07176888651318021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,64,0.035679999325010515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,32,0.016535111599498324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,1536,0.07233777973386976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,128,0.01901333365175459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,32,0.036369777388042875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,128,0.02348622183005015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,512,0.02671555512481266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,65536,0.9107191297743055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,1536,0.08169866932762994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,128,0.035338666703965925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,51200,0.7122844590081109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,65536,2.427290598551432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,16384,0.8463377952575684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,16384,0.6491093105740017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,51200,2.274740431043837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,12288,0.15674132770962185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,12288,0.5278035799662272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,10240,0.4290328820546468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,12288,0.4846293131510417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,10240,0.13991199599372015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,51200,2.6101440853542752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,65536,3.433968014187283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,10240,0.3869786792331272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,8192,0.10542133119371201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,16384,0.20509510570102266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,7168,0.2751493453979492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,7168,0.09529333644443089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,6144,0.23639289538065592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,6144,0.08744177553388809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,7168,0.267177767223782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,6144,0.23325067096286348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,4096,0.15057244565751818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,5120,0.19434844122992623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,4096,0.06092444393369886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,8192,0.34781421555413145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,3584,0.05496711201137967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,8192,0.30821066432529026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,3584,0.13553067048390707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,5120,0.20558133390214708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,2560,0.09664622280332778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,3072,0.04854489035076565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,5120,0.07148889038297865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,2048,0.08191911379496257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,2560,0.042790220843421094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,1536,0.061186664634280734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,2048,0.03719822234577603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,4096,0.15569155746036106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,2048,0.08445155620574951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,1536,0.030863109562132094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,1536,0.06725777520073785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,768,0.03196533189879523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,1024,0.05085600084728665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,3072,0.11924800607893203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,768,0.04121955566936069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,2560,0.10205866893132527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,3584,0.13393244478437635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,256,0.0196426659822464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,512,0.02107111116250356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,256,0.019307555423842538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,128,0.015847999188635085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,768,0.02306933369901445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,512,0.025969778498013813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,64,0.013539555172125498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,128,0.0192266669538286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,32,0.013644444445768992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,1024,0.04452088806364271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,64,0.028224001328150432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,1024,0.02546666728125678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,128,0.02812533246146308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,512,0.0337759984864129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,65536,0.9256578021579318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,3072,0.11738577153947617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,65536,2.159266577826606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,16384,0.5391350852118598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,51200,1.5417760213216145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,51200,0.7672266430324979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,16384,0.22689421971638998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,12288,0.3932906786600749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,256,0.02972888946533203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,16384,0.609240902794732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,12288,0.14821333355373806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,65536,3.0157928466796875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,10240,0.3281315432654487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,10240,0.12236000431908502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,8192,0.27547910478379994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,12288,0.5181582238939073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,8192,0.10754577981101142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,7168,0.23604355918036568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,51200,2.4229208628336587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,8192,0.28716354899936253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,6144,0.21889244185553658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,5120,0.15964089499579534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,32,0.027843554814656574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,6144,0.21850132942199707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,5120,0.06618311007817586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,4096,0.12673955493503147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,3584,0.11178221967485215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,4096,0.13986310693952772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,7168,0.0898817777633667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,3584,0.12093867195977105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,3072,0.09954755836062962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,7168,0.23769778675503203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,2560,0.09130044115914239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,3072,0.04475555486149258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,3072,0.10597511132558186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,2048,0.06459289126925997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,4096,0.05537777807977465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,2560,0.09079022539986505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,10240,0.33975733651055234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,2048,0.03311466508441501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,2048,0.07565777831607394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,1536,0.027767111857732136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,1024,0.038498666551378034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,1536,0.06126488579644097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,1024,0.02387733260790507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,1024,0.045238223340776235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,768,0.02863111098607381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,2560,0.03905244337187873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,512,0.02090311050415039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,768,0.03711644477314419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,512,0.018974221414989896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,512,0.030468444029490154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,6144,0.07865689198176067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,256,0.017492444978819955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,5120,0.17166845003763834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,768,0.021314667330847845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,1536,0.05263466636339823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,128,0.025412445267041523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,128,0.017629333668284945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,32,0.0125937776433097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,3584,0.050360000795788236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,256,0.017297777864668105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,64,0.02515999972820282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,32,0.024887111451890733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,256,0.026478222674793665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,128,0.014504889647165934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,64,0.012410666379663678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,65536,0.7359066539340549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,51200,0.6023564338684082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,16384,0.22285244199964735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,51200,1.4020995034111872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,65536,2.075927946302626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,12288,0.12501333819495306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,12288,0.3809804386562771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,10240,0.30167023340861004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,51200,2.1344178517659507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,10240,0.10677333010567559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,8192,0.23137066099378797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,10240,0.29535643259684247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,8192,0.08609866433673435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,12288,0.41054667366875547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,8192,0.24991289774576822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,7168,0.07692178090413411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,6144,0.18302755885654023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,7168,0.20651911364661324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,6144,0.06798399819268121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,5120,0.15332089530097112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,6144,0.1782897843254937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,4096,0.1102657781706916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,5120,0.058371557129753955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,65536,2.5819778442382812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,3584,0.10386488835016887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,4096,0.05127822359402975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,7168,0.1982497771581014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,3584,0.04292533463901944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,16384,0.491105768415663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,2560,0.0705804427464803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,3072,0.09370311101277669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,16384,0.4560231102837457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,2048,0.05508088734414843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,2560,0.0797555579079522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,2048,0.029607110553317603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,1536,0.04386577672428555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,3072,0.09491644303003947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,3584,0.1068906651602851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,1024,0.030781331989500258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,1536,0.05346488952636719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,2560,0.034662223524517484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,768,0.02390755547417535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,5120,0.1497244437535604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,2048,0.06640355454550849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,768,0.03275466627544827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,768,0.018924444913864136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,1536,0.02511466708448198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,256,0.014163555370436775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,512,0.026991999811596338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,256,0.015917332635985482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,128,0.01236088905069563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,4096,0.11994844012790257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,128,0.022475555539131165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,64,0.010563555690977307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,128,0.016215110818545025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,512,0.017285333739386666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,64,0.02233422299226125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,32,0.02206666602028741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,1024,0.040114667680528425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,512,0.018447111050287884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,3072,0.038482666015625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,32,0.010724444356229572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,65536,0.6479351255628798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,256,0.023053333163261414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,16384,0.4312177764044867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,51200,1.1380720138549805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,16384,0.14958400196499294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,16384,0.4180808862050374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,12288,0.30292977227105033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,65536,2.108393775092231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,12288,0.11209511756896973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,1024,0.021249777740902368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,10240,0.09512799978256226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,65536,1.7765946918063695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,8192,0.19607910845014784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,8192,0.0750782224867079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,8192,0.20113778114318848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,10240,0.21464710765414766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,7168,0.17719733715057373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,12288,0.31294133928087026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,7168,0.06846311357286242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,6144,0.16106311480204263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,10240,0.2568346659342448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,6144,0.05985155370500353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,7168,0.17658844259050158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,5120,0.05142133434613546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,4096,0.10430666473176743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,5120,0.12872800562116835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,4096,0.04225866662131416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,4096,0.10406933228174846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,51200,1.8337075975206163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,3584,0.03789600067668491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,51200,0.4958302179972331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,5120,0.1272453334596422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,3072,0.03421333432197571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,6144,0.15464177396562365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,3072,0.08040533463160197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,2048,0.04813511172930399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,2560,0.030214223596784804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,1536,0.041201776928371854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,3584,0.07868355512619019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,2048,0.025900445050663416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,1024,0.02678933408525255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,1536,0.022275555464956496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,1536,0.04598844382497999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,3584,0.09182222021950616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,768,0.02088533341884613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,1024,0.03457244568400913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,512,0.016920000314712524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,2560,0.06774399677912395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,768,0.028431998358832464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,256,0.012983111043771109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,512,0.023561777340041265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,2560,0.06964266962475248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,128,0.010986666712496014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,256,0.014704888065656027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,256,0.02029600077205234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,64,0.009409777820110321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,128,0.014803555276658801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,1024,0.018615111708641052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,768,0.016992888516849942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,32,0.019279999865425956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,2048,0.057387557294633657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,3072,0.06409333149592082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,128,0.0196106665664249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,65536,1.0721057256062825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,64,0.01943733294804891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,65536,0.5371822251213921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,512,0.01572533282968733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,16384,0.3050746652815077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,51200,0.8467520078023275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,16384,0.12239289283752441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,12288,0.23563377062479654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,16384,0.3601173294915094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,12288,0.09411555528640747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,32,0.009748444788985783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,12288,0.26729244656032985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,10240,0.08001955350240071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,10240,0.2165306674109565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,8192,0.06364799870385064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,7168,0.14921422799428305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,51200,0.4213111135694716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,7168,0.05729777945412529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,6144,0.11250221729278564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,7168,0.1478782229953342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,65536,1.7383911344740126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,10240,0.19948177867465547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,5120,0.09822044769922893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,4096,0.08104888598124187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,5120,0.042400889926486544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,8192,0.1631928947236803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,5120,0.10780355665418838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,4096,0.03517422080039978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,4096,0.08600355519188775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,3584,0.03135733471976386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,3584,0.07694755660163032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,3072,0.028952889972262915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,3072,0.06713155243131849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,2560,0.04649688800175985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,6144,0.128748443391588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,51200,1.297785758972168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,2560,0.026132444540659588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,3584,0.06848266389634874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,2560,0.05758488840527005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,1536,0.03107200066248576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,2048,0.022487110561794702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,8192,0.16818755202823213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,1024,0.020991111795107525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,2048,0.04919822348488701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,3072,0.05941422118080986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,768,0.017887999614079792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,1536,0.039156443542904325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,512,0.014117333624098035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,768,0.015298666225539314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,768,0.02442577812406752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,256,0.010463111102581024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,512,0.014178666803571912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,512,0.02017333275742001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,2048,0.04207999838723076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,128,0.00906488878859414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,256,0.01333866682317522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,256,0.0174346665541331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,64,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,128,0.013274666335847644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,128,0.016519111063745286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,32,0.008310221963458592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,1024,0.016203555795881484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,64,0.01647911138004727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,32,0.01642400026321411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,65536,0.44273776478237575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,1536,0.018912000788582694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,65536,1.2108728620741103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,1024,0.029331554969151814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,16384,0.3353946738772922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,51200,0.353312889734904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,65536,1.4989884694417317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,16384,0.13115821944342718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,16384,0.30294844839307994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,12288,0.10507288906309341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,12288,0.22884178161621094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,51200,1.208806249830458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,8192,0.1460720035764906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,10240,0.18723200427161324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,6144,0.04981333348486158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,7168,0.12924888398912218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,12288,0.2039759953816732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,7168,0.1341680023405287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,10240,0.19282755586836073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,10240,0.07179466883341472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,5120,0.09105688995785183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,8192,0.05640266338984171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,8192,0.15807910760243735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,4096,0.06958399878607856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,7168,0.05072888731956482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,5120,0.09812177552117242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,6144,0.10899111297395493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,4096,0.031606223848130964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,4096,0.07896977663040161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,3072,0.04717777834998237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,3584,0.07030755281448364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,3072,0.026178666286998328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,3072,0.06118844615088569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,2560,0.04169244567553202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,5120,0.03806133402718438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,2048,0.033815112378862165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,2560,0.023634667197863262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,3584,0.06592266427146064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,1536,0.028004444307751123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,51200,0.708821349673801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,2048,0.044405334525638156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,1024,0.02019644445843167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,1536,0.018145778112941317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,1536,0.0354017780886756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,768,0.015585778488053216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,1024,0.015356444650226168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,1024,0.02630311085118188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,768,0.014370666609870063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,512,0.012104888757069906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,768,0.021884444687101576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,256,0.009511999785900116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,2560,0.052652445104387074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,2048,0.020475554797384474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,128,0.008367111285527548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,3584,0.028569777806599934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,6144,0.11542488469017877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,64,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,128,0.012829333543777466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,128,0.015206222732861837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,32,0.007828444242477417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,64,0.014875556031862894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,32,0.015606222881211175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,512,0.01351111133893331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,512,0.018197332819302876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,256,0.012533333566453723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,6144,0.043741332160102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,65536,1.0728577507866752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,16384,0.25528889232211643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,51200,0.3239155610402425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,51200,0.6720648871527778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,16384,0.09824444188012017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,12288,0.19539199935065377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,16384,0.28278311093648273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,51200,0.9955093595716688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,65536,1.2244906955295138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,12288,0.21320533752441406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,8192,0.12800710731082493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,10240,0.16705689165327284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,8192,0.05180177754826016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,7168,0.12300622463226318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,7168,0.04580977890226576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,7168,0.11971822049882676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,6144,0.1034266683790419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,10240,0.16344977749718562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,6144,0.03979555434650845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,5120,0.08459022310045029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,10240,0.06606133116616143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,6144,0.10309688912497626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,65536,0.41059109899732804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,5120,0.0868684450785319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,4096,0.028509332074059382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,3584,0.05705955293443468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,12288,0.07650311125649346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,256,0.01591911084122128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,3584,0.025610667135980394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,3072,0.050069332122802734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,4096,0.06624977456198798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,2560,0.04015288750330607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,8192,0.13351821899414062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,3072,0.05438933438724942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,2048,0.03298755486806234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,2560,0.04674399892489115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,2048,0.018582221534517076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,1536,0.023997333314683702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,3584,0.061675555176205106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,5120,0.03460888730155097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,1024,0.01702666613790724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,1536,0.01642400026321411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,1536,0.03204355637232462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,768,0.014532445205582513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,1024,0.014255111416180929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,512,0.011794666449228922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,768,0.01977333260907067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,2048,0.03992444607946608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,256,0.009000889129108852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,256,0.014300444059901767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,256,0.0122079998254776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,128,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,512,0.012473777764373355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,128,0.013735999663670858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,64,0.007305777735180325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,3072,0.023023999399609033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,32,0.007340444458855524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,64,0.01370933320787218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,4096,0.07042933172649808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,32,0.013963555296262106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,512,0.016584888100624084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,65536,0.6604373190138075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,128,0.011805333197116852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,65536,0.363306654824151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,51200,0.5754319826761881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,51200,0.2796391116248237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,65536,1.0306310653686523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,16384,0.2404328982035319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,51200,0.8049137857225207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,12288,0.068904889954461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,12288,0.1749031146367391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,10240,0.1234640015496148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,16384,0.22042132748497856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,8192,0.109951999452379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,16384,0.0869813362757365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,8192,0.11745511160956489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,7168,0.09119288788901435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,2560,0.021192888418833416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,7168,0.10325422551896836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,10240,0.14742044607798258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,6144,0.07552978065278795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,12288,0.15862400001949734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,5120,0.06677777899636163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,6144,0.09012889199786717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,8192,0.047305779324637524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,4096,0.04665866825315687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,5120,0.03179377648565505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,5120,0.07703910933600532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,3584,0.04267377654711405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,768,0.013520888984203339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,4096,0.061384889814588756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,3072,0.03492177857293023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,3584,0.02459466622935401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,3584,0.05423733260896471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,2560,0.0328400002585517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,3072,0.022348443667093914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,3072,0.04794311192300585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,2048,0.025774222281244066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,2560,0.020114666885799833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,2560,0.041703111595577665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,1536,0.02065955599149068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,4096,0.026716444227430556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,1024,0.023927999867333308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,1024,0.015318221516079374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,1536,0.01514755520555708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,1536,0.028543998797734577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,768,0.012420444852775998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,1024,0.013212444053755866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,1024,0.021314667330847845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,512,0.010021333065297868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,768,0.01793866687350803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,512,0.012087999946541257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,512,0.015058666467666626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,2048,0.035460445615980364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,10240,0.05946311023500231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,128,0.0070062221752272705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,256,0.011400889191362592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,7168,0.0421635541650984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,256,0.012834666503800286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,64,0.006260444306664997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,128,0.01130222198035982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,32,0.006657777975002925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,128,0.012444444828563266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,768,0.012551999754375882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,64,0.012095111111799875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,32,0.01242844429281023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,2048,0.01758577757411533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,65536,0.49596712324354386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,65536,0.3015475538041857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,51200,0.4053226576911078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,16384,0.14723643991682264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,51200,0.23971199989318848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,65536,0.8291689025031196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,16384,0.07664622200859918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,51200,0.6435848871866862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,12288,0.06150311231613159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,10240,0.10344800021913315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,12288,0.14914933840433756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,256,0.007346666521496243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,10240,0.052128000391854175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,10240,0.12496356169382732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,8192,0.0403386652469635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,7168,0.07157778077655368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,8192,0.10143378045823838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,6144,0.0597662197219001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,12288,0.1127084493637085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,16384,0.1966844399770101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,6144,0.032003555032942034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,5120,0.04717866579691569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,6144,0.07771999968422784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,5120,0.028079999817742243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,8192,0.08029155598746406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,5120,0.0654239985677931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,3584,0.03347555465168423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,4096,0.0525511105855306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,3584,0.0215013325214386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,3584,0.04721688893106249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,6144,0.03657866517702738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,7168,0.03672088848219977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,2560,0.025681777132882014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,3072,0.019560888409614563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,7168,0.08817333645290798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,2048,0.021554667088720534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,2560,0.01805688937505086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,1536,0.0174026671383116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,2048,0.016221332881185744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,2048,0.03069066670205858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,1536,0.014534221755133735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,1024,0.013095999757448832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,1536,0.02482755482196808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,768,0.011053333679835001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,3072,0.030841777722040813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,1024,0.018742221924993727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,512,0.009109333157539368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,768,0.015855999456511605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,3072,0.041674666934543185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,256,0.008314666648705801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,512,0.011414222419261932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,512,0.013321778012646569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,128,0.005921777751710679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,256,0.01070844464831882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,256,0.011443555355072021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,64,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,128,0.010712888505723743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,128,0.01072088877360026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,64,0.01074488874938753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,32,0.010821333361996545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,4096,0.03707111212942336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,65536,0.43007556597391766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,4096,0.02397955622937944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,65536,0.24708978335062662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,51200,0.34623999065823025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,32,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,65536,0.6655662324693468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,768,0.011776000261306763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,51200,0.19498844941457114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,16384,0.07199733124838935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,12288,0.09555288818147446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,51200,0.5231733322143555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,10240,0.08105778031879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,12288,0.055381334490246244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,8192,0.0630177789264255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,12288,0.12424977620442708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,10240,0.10597155491511027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,7168,0.054802666107813515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,1024,0.012855110896958245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,8192,0.08433333370420669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,16384,0.11886221832699245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,6144,0.04652355445755852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,7168,0.031034668286641438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,5120,0.040587554375330605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,6144,0.02795733345879449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,16384,0.1652337842517429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,6144,0.06443022357092963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,4096,0.03203999996185303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,5120,0.024107555548350017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,3584,0.02678844498263465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,5120,0.05423288875155979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,10240,0.0458808872434828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,4096,0.020477333002620272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,4096,0.04492177897029453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,3584,0.01904266741540697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,3072,0.017297777864668105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,3072,0.035543110635545515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,2560,0.020047111643685233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,2560,0.03670400049951341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,2048,0.01697422232892778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,2560,0.03071911136309306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,2048,0.014865777558750577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,3072,0.023560000790490046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,3584,0.04044355617629157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,1536,0.01293155550956726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,1536,0.021391999390390184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,1024,0.010059555371602377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,2560,0.016018667154841952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,768,0.008454221818182204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,8192,0.034464889102511935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,512,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,768,0.011050666371981302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,7168,0.07418400049209595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,512,0.010448888772063784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,512,0.011619555453459421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,1024,0.011615999870830111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,1024,0.016199111938476562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,256,0.009869332942697737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,256,0.010057777994208867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,128,0.009699555734793345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,1536,0.01350844403107961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,128,0.009375111096435124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,32,0.004967111266321606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,64,0.009376889301670922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,32,0.009347555538018545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,65536,0.3444320095909967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,256,0.006235555642180973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,51200,0.26362400584750706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,65536,0.5197546746995714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,2048,0.026138666603300307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,16384,0.132189326816135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,16384,0.054267555475234985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,128,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,12288,0.0633680025736491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,64,0.005064888960785336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,51200,0.17550222078959146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,12288,0.10070400105582343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,10240,0.037547555234697126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,10240,0.08500444226794773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,8192,0.04277688927120633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,8192,0.027823110421498615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,7168,0.03611466619703505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,768,0.013844444519943662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,7168,0.05990755558013916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,7168,0.025270221961869135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,6144,0.029245333539115057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,6144,0.022251556317011516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,5120,0.024471110767788355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,6144,0.052150222990247935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,65536,0.19533422258165148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,8192,0.0680017802450392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,5120,0.019511111908488803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,5120,0.044418666097852916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,51200,0.4060542318556044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,4096,0.03679910964436001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,3584,0.015844444433848064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,3584,0.033357332150141396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,3072,0.016345777445369296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,3072,0.014510222607188754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,2560,0.014564444621404013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,4096,0.020941333638297185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,2560,0.013455111119482251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,16384,0.09004355801476373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,2048,0.012090666426552666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,2048,0.0123848890264829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,2048,0.02161600026819441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,3072,0.029269334342744615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,1536,0.011401777466138205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,1024,0.00868000007337994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,2560,0.02533600065443251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,1024,0.010446222292052375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,768,0.007287110719415877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,768,0.011843555503421359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,768,0.010056888891590966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,512,0.006048888795905643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,12288,0.043102221356497876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,512,0.009365333451165093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,256,0.0052800000541739995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,512,0.010194666683673859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,256,0.00925688859489229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,128,0.004560889055331548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,256,0.008691555923885768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,128,0.009001777403884465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,64,0.004223111189073986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,1024,0.013801777528391944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,32,0.00423200014564726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,1536,0.017794667018784415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,32,0.00776800016562144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,65536,0.24403201209174263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,128,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,51200,0.2100399997499254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,65536,0.19340622425079346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,1536,0.010328888893127441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,16384,0.06529866986804538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,51200,0.15336533387502035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,51200,0.35729511578877765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,16384,0.11368000507354736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,16384,0.05318311187956068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,12288,0.054089777999454074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,65536,0.4369688828786214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,10240,0.04211733407444424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,10240,0.07203288873036702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,10240,0.054619554016325206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,10240,0.03657777773009406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,8192,0.03279733326699998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,8192,0.027426666683620874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,64,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,7168,0.05156800150871277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,6144,0.023781332704755995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,4096,0.01706044375896454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,6144,0.021959111094474792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,6144,0.04521155688497755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,12288,0.041690667470296226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,8192,0.05818844503826565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,12288,0.08692444695366754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,4096,0.016912000046836007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,5120,0.01944266590807173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,5120,0.038132445679770574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,4096,0.016916443904240925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,4096,0.031587554348839655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,3072,0.013432888521088494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,3584,0.028241776757770117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,3072,0.014481777946154276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,3072,0.025271111064487036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,7168,0.02862488892343309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,2560,0.013510222236315409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,2048,0.01037599974208408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,7168,0.02424888809521993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,1536,0.008557333714432186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,2048,0.018983110785484314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,2048,0.012265777422322167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,1024,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,2560,0.011763555308183035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,1536,0.01571377780702379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,768,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,1024,0.012215110990736218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,2560,0.021934222843911912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,3584,0.014915555715560913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,512,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,768,0.010055111514197456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,256,0.004572444491916233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,512,0.009635555247465769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,512,0.009277332988050249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,128,0.003926222109132343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,256,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,5120,0.020868443780475195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,128,0.00904622260067198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,128,0.007154666715198093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,1024,0.010404444403118556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,64,0.0070062221752272705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,768,0.010411555568377176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,3584,0.015853333804342482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,65536,0.21668977207607695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,256,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,51200,0.17609688970777723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,1536,0.01146044416560067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,64,0.00425866660144594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,16384,0.057413332992129855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,51200,0.11819467279646133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,32,0.00388711111413108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,12288,0.043987555636299976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,51200,0.3083297676510281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,16384,0.04142133394877116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,10240,0.03535022338231405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,16384,0.10165511237250434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,32,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,12288,0.07658044497172038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,10240,0.029527998632854883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,7168,0.02407377792729272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,10240,0.06436800294452243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,8192,0.02045866681469811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,6144,0.0206666671567493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,7168,0.01885955532391866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,7168,0.046237332953347095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,5120,0.018274666534529794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,6144,0.040694223509894475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,12288,0.033013333876927696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,5120,0.03427911135885451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,8192,0.027456000447273254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,4096,0.013502221968438892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,65536,0.14708977275424534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,65536,0.385425779554579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,6144,0.016905777984195285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,3584,0.02567111121283637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,3584,0.012582222620646158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,3072,0.011807999677128263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,2560,0.010928000013033548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,8192,0.05197688937187195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,2560,0.020057777563730877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,2048,0.009232000344329411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,2560,0.011400889191362592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,3072,0.012246222131782107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,2048,0.010750222537252637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,2048,0.017214222086800467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,5120,0.01517777807182736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,3072,0.023031110564867657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,1536,0.010051555931568146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,1536,0.014301333162519666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,3584,0.014093332820468478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,1024,0.009347555538018545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,768,0.0058577776783042485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,1024,0.011058666639857821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,512,0.005249777601824866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,768,0.009692444569534725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,512,0.008685333033402761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,256,0.004223111189073986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,1024,0.006613333192136552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,256,0.0083155557513237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,256,0.007003555695215861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,128,0.004242666479614046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,128,0.008352888955010308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,768,0.009053332938088311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,64,0.003552888830502828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,1536,0.008113777471913232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,32,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,65536,0.12892799907260472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,4096,0.014846222268210517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,32,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,51200,0.10121066702736749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,4096,0.028287112712860107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,16384,0.039839999543295965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,51200,0.11706577406989203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,51200,0.2436906761593289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,16384,0.04092977775467767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,16384,0.0832115544213189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,512,0.008404444489214156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,12288,0.03271822134653727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,12288,0.06173688835567898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,65536,0.14572532971700033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,10240,0.028660443094041612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,10240,0.05252266592449612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,65536,0.3113120132022434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,7168,0.015261333849694995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,8192,0.020044444335831534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,128,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,8192,0.042103999190860324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,6144,0.01332711097266939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,7168,0.018530666828155518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,7168,0.03779911001523336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,5120,0.012139555480745105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,6144,0.016883555385801528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,10240,0.020429333051045735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,6144,0.033089776833852134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,5120,0.02860800094074673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,4096,0.013510222236315409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,4096,0.023312888211674158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,3584,0.012816000315878125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,3584,0.021254221598307293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,3072,0.008692444198661381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,12288,0.025412445267041523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,2560,0.007763555480374231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,3072,0.011765333513418833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,3072,0.019038221902317472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,64,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,2560,0.01142133358452055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,2560,0.01694222291310628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,1536,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,3584,0.009387555221716562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,8192,0.016626665989557903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,1024,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,1536,0.010016889207892949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,1536,0.012096888489193387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,768,0.004619555754794015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,1024,0.009390222529570261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,1024,0.009704888694816166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,512,0.004224888980388641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,768,0.009011555876996782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,4096,0.01053600013256073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,2048,0.006759111252095964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,256,0.0036951113078329298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,512,0.007355555891990662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,2048,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,128,0.003900444342030419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,2048,0.014527999692493014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,256,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,64,0.003560888684458203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,128,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,128,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,65536,0.08392355839411418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,32,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,768,0.008664888640244802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,5120,0.015206222732861837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,512,0.008399110701349046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,51200,0.06808711422814263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,65536,0.269815100563897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,256,0.008211555580298105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,51200,0.11667822466956244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,12288,0.01942044496536255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,51200,0.2158799966176351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,16384,0.07031200329462688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,10240,0.01771377854877048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,12288,0.05389155613051521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,10240,0.02940888868437873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,65536,0.1451315614912245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,8192,0.01309511148267322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,16384,0.028821332587136164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,8192,0.020928000410397846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,16384,0.041484445333480835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,7168,0.012089778151777057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,64,0.005665777872006099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,6144,0.010762666662534079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,7168,0.032772445016437106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,6144,0.017647999856207106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,5120,0.009393778112199571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,6144,0.028853333658642236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,5120,0.01570044457912445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,4096,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,5120,0.02476444509294298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,8192,0.03678222166167365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,3584,0.0075137780772315125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,4096,0.013841778039932251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,7168,0.019541333119074505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,3072,0.0069875551594628235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,3584,0.012826667063766055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,3584,0.018749333090252347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,10240,0.04525244567129347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,3072,0.012129777835475074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,3072,0.016979555288950603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,2048,0.005595555735958948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,2560,0.011400889191362592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,2560,0.015170666906568738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,32,0.003735110991530948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,2048,0.010769777827792697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,2048,0.013175999952687157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,1024,0.0042239998777707415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,1536,0.010082667072614035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,768,0.0038577777643998465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,1536,0.011122666299343109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,1024,0.009383999639087254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,512,0.003563555578390757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,1024,0.009012444151772393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,768,0.009053332938088311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,256,0.0032222221295038858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,512,0.008622222476535374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,512,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,128,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,256,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,128,0.008073777788215214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,12288,0.033598221010631986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,64,0.002874666617976295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,4096,0.020863110820452373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,2560,0.006263999889294307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,768,0.008010666403505538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,32,0.005349333501524395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,51200,0.0661884413825141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,256,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,1536,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,16384,0.025125333004527624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,12288,0.01499644418557485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,128,0.0052782222628593445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,51200,0.2105635537041558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,10240,0.013457777599493662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,16384,0.07014755407969157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,8192,0.01220533334546619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,7168,0.01144177797767851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,12288,0.053824888335333936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,6144,0.010384000009960597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,65536,0.08094933297899035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,8192,0.03648444347911411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,5120,0.009010666774378883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,4096,0.007975999679830339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,6144,0.028548445966508653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,3584,0.007287999822033777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,4096,0.020593777298927307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,3072,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,32,0.0029075555503368378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,2560,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,64,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,2048,0.0052373334765434265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,2560,0.015096889601813423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,1536,0.004592000196377437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,2048,0.013156444662147097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,1024,0.003919110943873723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,1536,0.011120888921949597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,768,0.003896000070704354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,1024,0.008893333375453949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,512,0.003552000141806073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,3072,0.01688799924320645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,256,0.003192000091075897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,512,0.0069822221994400024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,128,0.0028479999552170434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,256,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,5120,0.024451555477248296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,64,0.0028702221396896574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,32,0.0028524444335036804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,128,0.00526577772365676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,65536,0.07544266515307956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,32,0.0052888890107472735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,10240,0.04542488853136698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,7168,0.032567110326555036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,65536,0.27508266766866046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,16384,0.02394666605525547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,12288,0.012800888882742988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,65536,0.2681271235148112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,51200,0.21027199427286783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,10240,0.011753777662913004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,16384,0.0719111098183526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,8192,0.010097777677906884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,64,0.0052728888889153796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,7168,0.010055999788973067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,10240,0.04533866710133023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,768,0.008012444608741337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,8192,0.036638223462634616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,5120,0.011436444189813403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,7168,0.032431999842325844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,4096,0.00980533328321245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,6144,0.028530667225519817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,3584,0.00902133352226681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,5120,0.024364444944593642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,3072,0.007967111137178209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,4096,0.020640888147883944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,3584,0.01891644464598762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,12288,0.05350755651791891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,3584,0.01924088928434584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,3072,0.017180444465743173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,51200,0.060759113894568555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,1024,0.00423911131090588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,2048,0.013376000026861826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,768,0.003879111260175705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,1536,0.011076444553004371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,768,0.008034666379292807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,2560,0.007328888608349695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,2048,0.005296000176005893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,512,0.007003555695215861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,128,0.0028560000161329904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,256,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,64,0.0028479999552170434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,128,0.005295111073387994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,1024,0.009026666482289633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,512,0.0035377778112888336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,256,0.0032115555885765287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,6144,0.00908799966176351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,2560,0.01516711049609714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,32,0.002895999906791581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,64,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,32,0.005048888838953442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,1536,0.004575110971927643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,16384,1.4321058061387804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,16384,4.573440975613065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,16384,5.075711144341363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,51200,6.007825639512803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,12288,3.794879065619575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,16384,4.498578813340929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,12288,1.0925893783569336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,12288,3.5299546983506946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,10240,0.9176435470581055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,10240,3.345106760660807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,10240,3.0163777669270835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,12288,3.239312065972222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,8192,0.6955822308858236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,10240,2.727657741970486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,8192,2.7228800455729165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,51200,14.183141072591146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,8192,2.193129857381185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,7168,2.1479803721110025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,8192,2.246982150607639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,7168,0.6261093351576064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,7168,1.98797607421875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,6144,1.6407164467705622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,6144,0.5467066764831543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,6144,1.6717679765489366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,6144,1.6381848653157551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,5120,0.47205066680908203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,5120,1.382030275132921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,7168,1.8970027499728734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,5120,1.3808195326063368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,4096,1.2862168418036568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,4096,1.124283578660753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,3584,1.0554516050550673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,4096,1.076497819688585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,5120,1.4105653762817383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,3584,0.39557245042588973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,3584,0.9455039766099719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,3072,0.916752921210395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,3072,0.8310942120022244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,3072,0.3015493286980523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,4096,0.4010382228427463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,2560,0.7371777958340116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,3072,0.7818426556057401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,51200,15.852873060438368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,2560,0.2746960057152642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,51200,21.847235785590275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,2560,0.7045946651034884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,2048,0.23283200793796113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,2560,0.6551040013631185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,2048,0.6267395549350315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,1536,0.4823297924465603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,3584,0.9674284193250867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,1536,0.18732533189985487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,1024,0.2822444438934326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,1536,0.4543742073906793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,768,0.2362133396996392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,2048,0.5614737934536403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,1024,0.319160885281033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,768,0.2688808970981174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,1024,0.47449424531724715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,512,0.17726843886905244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,2048,0.571157349480523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,512,0.22969423400031197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,512,0.11215555667877197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,512,0.21647377808888754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,256,0.2074497805701362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,768,0.2662755648295085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,128,0.09607822365230984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,256,0.09903022315767075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,1536,0.4325839943355984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,64,0.08837688631481594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,128,0.09744889206356472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,128,0.12570133474138048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,128,0.19535199801127115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,32,0.09226044681337144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,768,0.1295484436882867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,64,0.1994702286190457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,32,0.20214400026533338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,1024,0.1479884386062622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,256,0.16516355673472086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,256,0.11336444483862983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,65536,5.762081570095486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,51200,4.279977586534288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,16384,3.9849387274848094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,51200,10.980423821343315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,16384,4.350766075981988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,16384,1.1227475272284615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,65536,14.167549981011284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,12288,2.6377991570366754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,12288,0.8333822356330024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,16384,3.5591138203938804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,12288,2.9734747144911022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,51200,16.87134297688802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,12288,2.535890579223633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,51200,12.8663330078125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,10240,2.47508790757921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,10240,2.4779351552327475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,10240,2.1212141248914933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,10240,0.6823182106018066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,8192,0.5576275719536675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,8192,1.861496819390191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,65536,20.68237135145399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,65536,15.447040134006075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,8192,1.6719173855251734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,7168,1.621339586046007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,7168,1.5418062210083008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,6144,1.3720489078097875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,6144,0.4319217734866672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,7168,1.4992604785495336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,7168,0.4922435548570421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,5120,1.0257973141140408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,6144,1.3279422124226887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,6144,1.271026717291938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,5120,1.0845777723524306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,4096,0.9182524151272244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,8192,1.9243094126383464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,5120,1.039748403761122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,4096,0.8785386615329318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,3584,0.6681973139444987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,5120,0.3748195436265733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,3584,0.2777351008521186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,3584,0.8339706526862251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,3072,0.23984710375467935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,4096,0.3094826539357503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,3072,0.6723626454671224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,2560,0.5315155453152126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,4096,0.8225626415676541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,2560,0.21138755480448404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,2560,0.5480462180243598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,3072,0.589263121287028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,2560,0.5277866787380642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,3584,0.7158791224161783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,2048,0.18038577503628203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,1536,0.3197626537746853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,2048,0.4193342261844211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,3072,0.6248417960272895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,1536,0.14534933037228054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,1024,0.22230667538113066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,1536,0.3692595428890652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,1536,0.3292773299747043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,2048,0.4376853307088216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,1024,0.11608000596364339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,1024,0.2483688990275065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,1024,0.26982400152418345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,512,0.14717155032687718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,2048,0.44951735602484805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,768,0.21234756045871309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,768,0.10063466760847305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,256,0.10153333346048991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,512,0.1809626685248481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,512,0.1656551096174452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,128,0.07739022043016222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,256,0.07918044593599108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,256,0.13076089488135442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,64,0.06453955835766263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,128,0.07781777779261272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,128,0.1537991099887424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,32,0.06631377670500013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,128,0.10018489095899795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,64,0.15720710489484999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,512,0.08973955445819431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,32,0.15982400046454534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,256,0.1635040044784546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,768,0.17374489042494032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,768,0.20482756031884086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,65536,1.9039164649115667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,51200,1.394197358025445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,16384,1.3244444529215496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,51200,3.8253856235080295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,65536,5.013554467095269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,16384,0.367611567179362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,16384,1.3132283952501085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,12288,0.9835110770331489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,16384,1.1237209108140733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,51200,5.306913587782118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,12288,0.2946915626525879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,12288,0.784789350297716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,65536,5.0288268195258246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,65536,7.2090852525499125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,10240,0.2306471135881212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,10240,0.7595226499769422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,10240,0.6533582475450304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,8192,0.6368284225463867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,51200,3.7607243855794272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,8192,0.6054666837056478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,8192,0.5165333218044704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,7168,0.5382382074991862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,7168,0.17046844959259033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,7168,0.5190186500549316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,7168,0.44014933374192977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,12288,0.9430683983696831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,6144,0.14854578177134195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,8192,0.18974755869971383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,5120,0.38254488839043516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,10240,0.8294533093770345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,5120,0.1271244419945611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,5120,0.36054489347669816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,5120,0.3220906787448459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,6144,0.45744090610080296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,4096,0.2836035622490777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,3584,0.24689777692159018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,6144,0.46720091501871747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,6144,0.3835795455508762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,3584,0.0945884452925788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,3584,0.2391866577996148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,4096,0.2985031074947781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,3072,0.0867591102917989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,3072,0.21828356054094103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,4096,0.10776444276173909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,2560,0.07494044303894043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,2560,0.18579199579026964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,3584,0.24925867716471353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,3072,0.21334755420684814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,2048,0.06406488683488634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,1536,0.126310216055976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,2048,0.1448284387588501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,2560,0.1759493350982666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,1536,0.052095111873414784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,4096,0.2603030999501546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,1024,0.081968002849155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,1536,0.11771644486321343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,2048,0.15016267034742567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,768,0.07081511285569933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,1024,0.044066667556762695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,1024,0.09193688631057739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,2048,0.15329333146413168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,1024,0.08875644207000732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,512,0.054903109868367515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,768,0.03970311085383097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,768,0.07396977477603488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,3072,0.20266932911343047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,256,0.03966666592491998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,768,0.07497777541478474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,512,0.06322133541107178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,512,0.060927112897237144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,256,0.03228444523281521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,256,0.056998219754960805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,256,0.048135999176237315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,64,0.025411556164423626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,128,0.031889776388804116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,32,0.026808000273174707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,64,0.05446755554940966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,512,0.03603200117746989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,128,0.029695110188590154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,2560,0.172198216120402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,128,0.05344622333844503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,128,0.03877866599294875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,32,0.05526489019393921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,1536,0.12194844086964925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,65536,1.2958676020304363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,51200,0.9937537511189779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,65536,3.6280869377983938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,16384,1.049240854051378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,51200,2.9970525105794272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,16384,0.30425866444905597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,16384,1.027350213792589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,12288,0.7418906423780652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,16384,0.847815990447998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,12288,0.7018284267849393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,12288,0.21533511744605172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,65536,3.7155130174424915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,51200,2.914437399970161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,10240,0.5681431028578017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,10240,0.17892089154985216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,10240,0.5735431247287327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,8192,0.4877760145399306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,51200,4.250843471950955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,8192,0.16642844676971436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,12288,0.5892133182949489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,7168,0.3918737835354275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,8192,0.3949280050065782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,7168,0.13463377952575684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,7168,0.3923644489712185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,6144,0.35433512263827854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,7168,0.3446693420410156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,10240,0.4804657830132379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,6144,0.12098666032155354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,6144,0.33250223265753853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,5120,0.2876124382019043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,8192,0.4425013330247667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,5120,0.0997884472211202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,4096,0.19171199533674452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,5120,0.24622665511237252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,4096,0.08329866992102729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,4096,0.21790577305687797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,4096,0.20108266671498617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,3584,0.07523022095362346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,3584,0.1933822234471639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,3584,0.18172000514136422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,6144,0.2939537896050347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,5120,0.27320178349812824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,3072,0.06653422117233276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,2560,0.13330311245388454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,3584,0.1661964390012953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,2048,0.12352622879876031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,2560,0.05763644642300076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,2560,0.14321688810984293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,2560,0.13509778181711832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,65536,5.049265967475043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,2048,0.05044710967275831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,2048,0.1127600007587009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,3072,0.16797510782877603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,1536,0.09443910916646321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,1536,0.041783110962973706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,3072,0.1574017736646864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,1536,0.09162755807240804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,768,0.051947553952534996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,1024,0.03572977913750543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,1024,0.07048800256517199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,3072,0.16448711024390325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,512,0.044682668315039746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,768,0.05773599942525228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,2048,0.11874755223592122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,768,0.05882222122616238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,256,0.0336408880021837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,512,0.029495110114415485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,1024,0.06604799959394667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,512,0.04953333404329088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,512,0.04888711041874356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,256,0.026690666874249775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,256,0.044166223870383374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,256,0.03870933254559835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,64,0.02787999974356757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,128,0.03131022387080722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,1536,0.08414577775531346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,64,0.04231199953291151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,32,0.04305422306060791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,128,0.029887109994888306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,1024,0.07148977782991198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,128,0.026224889689021643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,768,0.032511999209721885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,32,0.028808888461854722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,128,0.04162044326464335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,65536,3.045141432020399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,51200,2.372464921739366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,51200,0.8451626565721301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,65536,1.126003583272298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,16384,0.7187875641716851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,16384,0.23846401108635795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,16384,0.813643561469184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,51200,2.4675964779324002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,16384,0.718632009294298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,65536,3.1350311703152127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,12288,0.5716222127278646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,12288,0.19798755645751953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,10240,0.473540465037028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,10240,0.15664177470737034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,10240,0.49342579311794704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,10240,0.4059680038028293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,12288,0.5074568854437934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,51200,3.662665049235026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,8192,0.3704240057203505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,7168,0.2918337715996636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,8192,0.3270417849222819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,65536,5.0807414584689665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,7168,0.32396621174282497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,12288,0.491504881117079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,6144,0.286847988764445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,7168,0.2920568784077962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,8192,0.435807122124566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,6144,0.09736444552739461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,6144,0.2796124352349175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,6144,0.2495502101050483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,8192,0.1242248879538642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,5120,0.0856871141327752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,4096,0.18339822027418348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,5120,0.23169777128431532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,5120,0.2103466722700331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,4096,0.18581777148776582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,4096,0.17427378230624727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,5120,0.23263022634718153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,3072,0.14742755889892578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,7168,0.11111199855804443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,3072,0.14223110675811768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,3584,0.14107288254631892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,2560,0.10221422380871242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,3584,0.16263911459181044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,3072,0.13669599427117243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,3584,0.15312355094485813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,2048,0.09243110815684001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,2560,0.04993333419164022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,2560,0.11608355575137669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,2048,0.10078577862845527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,1536,0.0735315548049079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,2048,0.09898844692442153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,3072,0.05719377597173055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,1536,0.03735911183887058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,1536,0.08066399892171223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,3584,0.06424355506896973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,1536,0.07948266797595553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,768,0.038169778055614896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,2560,0.12229511472913955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,1024,0.06193688842985365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,1024,0.061724444230397545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,512,0.029954665237002905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,768,0.028232001596026953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,768,0.049307554960250854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,768,0.05112444361050924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,4096,0.07015111049016316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,512,0.02529511186811659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,512,0.04179644584655762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,1024,0.051761776208877563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,128,0.01961599952644772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,256,0.03695644603835212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,256,0.033710221449534096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,128,0.03512533174620734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,128,0.022700443863868713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,64,0.01687111126052009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,32,0.016763556334707472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,256,0.022005332840813532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,64,0.035261332988739014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,32,0.03586133321126302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,256,0.023374223046832617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,1024,0.030990223089853924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,128,0.027339556150966223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,512,0.04348799917432997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,2048,0.04328088959058126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,65536,2.852831098768446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,65536,0.8516471121046277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,51200,2.2896595001220703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,51200,0.6782311333550347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,16384,0.7108284632364908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,16384,0.23303021325005424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,16384,0.6413120163811578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,51200,2.96525149875217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,16384,0.5805590947469076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,12288,0.5496542188856337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,10240,0.38995912339952254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,12288,0.5009493298000759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,12288,0.4020017782847087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,65536,3.4038488599989147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,10240,0.3776417838202582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,10240,0.4635964499579535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,51200,2.010942247178819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,8192,0.10500888692008124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,8192,0.27594221962822807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,7168,0.2778284549713135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,10240,0.12796444363064235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,8192,0.25919198989868164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,7168,0.24266576766967773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,6144,0.23311376571655273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,12288,0.14994133843315974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,6144,0.22913776503668892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,6144,0.2100835508770413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,7168,0.09170044130749172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,7168,0.26635999149746364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,5120,0.17685243818495008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,5120,0.06924000051286486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,8192,0.29886666933695477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,4096,0.15141422218746609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,5120,0.18723644150627983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,4096,0.05829955471886528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,5120,0.1910231113433838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,65536,2.5618373023139105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,3584,0.052720887793434985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,6144,0.08776711093054877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,3072,0.11791022618611653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,4096,0.14560533894432917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,3584,0.10164978106816609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,2560,0.07455378108554415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,3584,0.13433689541286892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,3072,0.10139289167192246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,2560,0.10073866446812947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,2560,0.042239109675089516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,2560,0.09750755627950032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,2048,0.061066667238871254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,2048,0.036729776197009616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,2048,0.08335110876295303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,1536,0.05173066589567396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,2048,0.08184978034761217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,3072,0.04811199837260776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,1024,0.038378665844599404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,1536,0.030115554730097454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,4096,0.1481457816229926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,1536,0.0667475528187222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,768,0.03179644544919332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,1024,0.051015999582078725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,1024,0.051980445782343544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,768,0.04044444362322489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,768,0.044292443328433566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,3072,0.11398666434817845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,3584,0.12828177875942653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,512,0.033477332856920027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,256,0.021390222840838965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,1024,0.024263999528355066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,256,0.018434666925006442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,128,0.016691555579503376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,768,0.022620444496472675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,512,0.02647022240691715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,128,0.028107557031843398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,64,0.01389600005414751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,128,0.023592889308929443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,32,0.015944888194402058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,512,0.03614755471547445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,64,0.02751733362674713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,256,0.029273778200149536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,256,0.028594666057162817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,128,0.018556444181336295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,1536,0.06653066476186116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,512,0.020255110330051847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,32,0.02749600013097127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,65536,2.303490744696723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,51200,0.6842240227593316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,51200,1.8470550113254125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,65536,0.8532133102416992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,65536,2.232886208428277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,16384,0.19000711705949572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,16384,0.6303457683987087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,51200,2.5545679728190103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,12288,0.4055146641201443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,12288,0.14246755176120335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,12288,0.3999137878417969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,16384,0.5296782387627496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,12288,0.354677332772149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,10240,0.11740533510843913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,8192,0.27348266707526314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,51200,1.7170222600301106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,10240,0.2919600009918213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,16384,0.5342888832092285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,8192,0.26946579085456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,8192,0.23937688933478463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,10240,0.32566489113701713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,7168,0.08466222551133896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,6144,0.20305866665310332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,7168,0.20794310834672716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,10240,0.33399200439453125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,6144,0.07354933023452759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,6144,0.2019066678153144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,8192,0.09418044487635295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,7168,0.2332328955332438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,4096,0.12965510951148138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,5120,0.15348266230689153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,7168,0.23367288377549914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,65536,3.2776737213134766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,3584,0.10905600256390041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,5120,0.18158843782213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,6144,0.18344888422224256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,3584,0.1202097733815511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,5120,0.06299022171232435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,3072,0.09731377495659722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,3584,0.1127671135796441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,3072,0.04236355423927307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,2560,0.08778755532370673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,4096,0.12675644291771784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,2560,0.0901635554101732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,3584,0.04815466536415947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,2560,0.08626044458813137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,2048,0.06644533077875774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,3072,0.1050880021519131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,2048,0.07446044683456421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,3072,0.10076533423529731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,1536,0.05276266733805338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,2048,0.03168711066246033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,1536,0.02678044471475813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,1536,0.05951733059353299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,1024,0.03658933440844218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,1536,0.059545775254567467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,2560,0.03712977634535895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,1024,0.044938666952980884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,1024,0.04646844334072537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,4096,0.1346542172961765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,2048,0.0747493306795756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,768,0.03649066554175483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,768,0.020003555549515616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,768,0.03899377915594313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,5120,0.17015911473168266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,512,0.030191110240088567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,512,0.01824355622132619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,512,0.032837334606382586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,768,0.02918133470747206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,4096,0.05401600069469876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,128,0.014167111780908374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,256,0.01711466742886437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,256,0.025207999679777358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,512,0.02036533256371816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,128,0.02493600050608317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,64,0.01516266663869222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,128,0.02167911165290409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,64,0.024784000383483037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,32,0.024725332856178284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,1024,0.022525333695941504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,256,0.025895110434956018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,128,0.017277333471510146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,32,0.015669332610236276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,65536,0.7424124611748589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,256,0.019519110520680744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,16384,0.49318668577406144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,51200,0.5885919994778103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,51200,1.787994596693251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,65536,2.3001030815972223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,16384,0.16944800482855904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,16384,0.5014062457614475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,65536,1.9644603729248047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,12288,0.41748089260525173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,65536,2.8694097730848522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,16384,0.46944443384806317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,12288,0.1300168832143148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,12288,0.36042043897840714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,12288,0.3180275493197971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,10240,0.10422933101654053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,8192,0.24350754419962564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,10240,0.2797813415527344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,8192,0.08397333489523993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,8192,0.23097154829237196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,8192,0.2173466682434082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,51200,2.0784852769639754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,51200,1.5262542300754125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,7168,0.07539466354582045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,6144,0.16301600138346353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,7168,0.18993243906233045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,6144,0.06535110870997111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,6144,0.17725777626037598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,6144,0.1642977794011434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,7168,0.20550044377644858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,5120,0.14769244194030762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,5120,0.13990755875905356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,10240,0.293832884894477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,4096,0.11852088239457871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,4096,0.04802044563823276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,3584,0.09287466605504353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,5120,0.13914488421546087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,3584,0.04351733459366692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,3584,0.10524711343977187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,3072,0.08048711220423381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,3584,0.10364177491929795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,4096,0.1199680036968655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,2560,0.07232444153891669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,3072,0.03856622179349264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,3072,0.09254133039050633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,3072,0.09305155277252197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,2048,0.057188444667392306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,4096,0.116230222913954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,2560,0.07924089166853163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,2560,0.07931022511588202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,1536,0.0432951119210985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,2048,0.06623911195331149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,5120,0.05719111363093058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,1536,0.024495111571417913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,1536,0.052998221582836576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,1024,0.031714667876561485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,1536,0.05550311009089152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,2560,0.03333866596221924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,10240,0.3109475506676568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,1024,0.020953777763578627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,2048,0.028459555572933618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,768,0.03253600001335144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,2048,0.06669777631759644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,512,0.018558222386572096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,768,0.03598755598068237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,7168,0.20563111040327284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,512,0.017384888397322763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,256,0.018565333551830716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,512,0.029782222376929388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,256,0.015912888778580558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,256,0.02311733365058899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,1024,0.042141334878073804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,768,0.019184889064894784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,128,0.019633778267436557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,512,0.0266311119000117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,64,0.015032000011867948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,768,0.024199111594094172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,128,0.016509332590632968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,64,0.02196177840232849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,128,0.02200088898340861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,32,0.02171733313136631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,128,0.016528000434239704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,1024,0.039805332819620766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,256,0.022882666852739122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,65536,1.355327076382107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,65536,0.602892451816135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,51200,1.0464435153537326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,51200,0.47221867243448895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,32,0.015615110596021017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,16384,0.3851582209269206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,65536,2.3265813191731772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,16384,0.14077422353956434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,12288,0.296341339747111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,16384,0.4406897756788466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,51200,1.8011031680636937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,16384,0.38446844948662656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,12288,0.30316800541347927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,10240,0.22593243916829428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,12288,0.2705315483940972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,10240,0.26245066854688853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,10240,0.22519556681315103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,65536,1.649687131245931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,7168,0.1747128831015693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,8192,0.19835289319356283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,8192,0.18137066894107393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,51200,1.2356612947252061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,12288,0.10574399762683445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,7168,0.0640373362435235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,6144,0.1433688931994968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,7168,0.16079378128051758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,10240,0.08795733584298028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,6144,0.056558224889967174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,5120,0.12005778153737386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,6144,0.15039643976423475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,6144,0.13929066393110487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,4096,0.09676800171534221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,5120,0.12676355573866102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,5120,0.11869422594706218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,4096,0.040764444404178195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,3584,0.08030666907628377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,4096,0.09830311271879409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,7168,0.17511733373006186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,3584,0.03606044583850437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,3584,0.09144088957044814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,3584,0.08799555566575791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,5120,0.04920977685186598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,3072,0.07958844635221693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,8192,0.0729057788848877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,3072,0.08000622193018596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,2560,0.02940888868437873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,2560,0.0683857798576355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,2048,0.04585244589381748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,2560,0.06780533658133613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,3072,0.07143288850784302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,2048,0.02518044412136078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,1536,0.03658488723966811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,2048,0.05689599778917101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,2048,0.05853422482808431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,1024,0.025271111064487036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,3072,0.03232977787653605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,1536,0.04571200079388089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,1536,0.046684443950653076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,768,0.01979377700222863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,1024,0.034854223330815635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,1024,0.03698755635155572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,768,0.028287112712860107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,8192,0.17341600524054634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,512,0.015239111251301236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,768,0.03118488854832119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,512,0.015494222442309061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,1536,0.021670222282409668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,2560,0.058825777636633984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,1024,0.018160000443458557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,4096,0.10303644339243571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,256,0.019967110620604623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,128,0.011045333411958484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,256,0.014194667339324951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,768,0.016906667086813185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,128,0.014493332968817817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,64,0.009691555466916826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,32,0.009905777871608734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,128,0.017960889471901786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,64,0.019168888529141743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,512,0.026176000634829204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,32,0.018953777021831937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,256,0.0199937770764033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,128,0.019305777218606737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,65536,0.4882044262356228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,65536,1.4562675688001845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,256,0.012060444388124677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,16384,0.32901154624091256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,51200,1.1534577475653756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,51200,0.38786665598551434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,65536,1.7756551106770833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,65536,1.3967101838853624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,16384,0.11694844563802083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,16384,0.33453067143758136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,51200,1.3326355616251628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,16384,0.3296017646789551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,51200,1.0357209311591256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,12288,0.09478488895628188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,12288,0.24959643681844076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,12288,0.23539643817477754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,10240,0.07738400167889066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,10240,0.2157315545611911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,8192,0.15675466590457493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,10240,0.19953778054979113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,8192,0.06391466988457574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,8192,0.16559555795457628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,512,0.023431999815834895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,10240,0.21417778068118623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,7168,0.14948533640967474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,7168,0.14101333088344997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,6144,0.12541600068410239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,5120,0.09403110875023736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,7168,0.1436177757051256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,8192,0.16166933377583823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,5120,0.10706044567955865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,5120,0.041259557008743286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,5120,0.10536444187164307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,6144,0.1266053385204739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,4096,0.03425777620739407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,4096,0.08595022228029038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,3584,0.07183111376232572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,4096,0.08732799688975017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,6144,0.122688889503479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,12288,0.2658204502529568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,3072,0.06115466356277466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,3584,0.03126577867401971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,3584,0.07627199755774604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,2560,0.051615112357669406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,3072,0.028132445282406274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,7168,0.055959999561309814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,3072,0.067193779680464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,4096,0.07366222143173218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,3072,0.06877599822150336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,2560,0.025262221693992615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,2560,0.05771555503209432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,2560,0.05985955397288004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,2048,0.021672889590263367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,2048,0.050438221957948476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,6144,0.04926933182610405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,1536,0.03888444436921014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,1536,0.04208711120817396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,3584,0.07805244127909343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,1024,0.029013332393434312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,768,0.01699466672208574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,1024,0.03254222207599216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,1536,0.029750221305423315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,2048,0.04821955495410495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,512,0.014224888549910652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,768,0.015090665883488126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,768,0.027407111393080816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,512,0.019861333900027804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,512,0.01404800017674764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,512,0.022915555371178523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,1024,0.01608800060219235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,256,0.017232000827789307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,128,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,768,0.0240346673462126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,1024,0.02100444502300686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,128,0.016355555918481614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,64,0.008653333617581261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,2048,0.03801244497299194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,32,0.00888177752494812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,256,0.013154666456911298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,256,0.01850755512714386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,32,0.016241777274343703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,1536,0.018758222460746765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,128,0.01328888866636488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,65536,0.44274221526251895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,64,0.016519111063745286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,51200,0.8823724322848849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,65536,1.461216926574707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,65536,1.2179288864135742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,256,0.01074488874938753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,16384,0.27587199211120605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,51200,0.34567822350396055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,65536,1.2161315282185872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,51200,1.188977771335178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,51200,0.9377386305067273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,16384,0.10374489095475938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,16384,0.2805039882659912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,12288,0.0827946662902832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,12288,0.22120089001125762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,12288,0.21357777383592394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,10240,0.07035466697480944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,10240,0.1870959997177124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,8192,0.14259555604722765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,10240,0.17719466156429717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,8192,0.0563102232085334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,8192,0.1482879983054267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,8192,0.14307378398047552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,16384,0.30851109822591144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,128,0.016685333516862657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,7168,0.1320799986521403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,7168,0.04993955625428093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,7168,0.12670844131045872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,6144,0.04331199990378486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,6144,0.11407289240095352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,5120,0.08546844455930923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,6144,0.11048444774415757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,7168,0.12382666269938152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,4096,0.0653653343518575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,12288,0.2032639980316162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,5120,0.09459466404385036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,10240,0.16356444358825684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,3584,0.05522133244408501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,4096,0.03146933184729682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,4096,0.08061422242058648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,3072,0.047693332036336265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,3584,0.06919466786914401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,3584,0.07036711110009088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,5120,0.0373671121067471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,3072,0.0607244438595242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,3072,0.02567111121283637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,3072,0.06290133131874932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,6144,0.09869333108266194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,2560,0.02330933345688714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,2048,0.033864001433054604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,2560,0.05439733465512594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,3584,0.028899557060665552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,2048,0.02053955528471205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,2048,0.04390844371583727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,2048,0.04643911123275757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,5120,0.09604089127646552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,1536,0.03544355432192484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,1024,0.019567999574873183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,1536,0.0381368895371755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,4096,0.07748533619774713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,2560,0.052300443251927696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,1024,0.02630222174856398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,1024,0.02933244572745429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,768,0.01425155500570933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,768,0.021976888179779053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,512,0.012415111064910889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,768,0.0251226673523585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,512,0.013270222478442721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,2560,0.04290489024586148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,256,0.0107360002067354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,512,0.021181333396169875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,256,0.012440889245933957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,768,0.015666666958067153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,128,0.009352888498041365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,256,0.016973333226309884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,1536,0.027544000082545813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,128,0.015214223000738354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,64,0.00868622213602066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,128,0.016255999604860943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,32,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,1536,0.018217777212460835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,64,0.014965333872371249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,32,0.01518311103185018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,1024,0.015728889240158927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,65536,0.39272001054551864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,65536,1.0721644295586479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,51200,0.6829226811726888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,512,0.018414222531848483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,65536,1.204246203104655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,65536,1.058831108940972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,256,0.01572800013754103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,51200,0.3133457766638862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,12288,0.1870417727364434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,128,0.012649777862760754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,51200,0.7769538031684028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,16384,0.2583555645412869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,12288,0.08148088720109728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,12288,0.19659199979570177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,12288,0.18391822444068062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,16384,0.2469111018710666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,10240,0.06344266732533772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,10240,0.15384088622199163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,16384,0.09559822082519531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,16384,0.25961777899000377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,8192,0.13296711444854736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,7168,0.11713688903384739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,10240,0.1618071132236057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,8192,0.12580266263749865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,6144,0.09149155351850723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,7168,0.11688621838887532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,7168,0.11274400022294785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,10240,0.16974222660064697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,5120,0.08151466978920831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,6144,0.10138755374484593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,51200,0.9880871242947049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,5120,0.03314133485158285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,5120,0.08646577596664429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,4096,0.06067644225226509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,5120,0.08380177948209976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,7168,0.04401333464516533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,4096,0.028075555960337322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,8192,0.13085777229732937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,6144,0.03872799873352051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,6144,0.0979306697845459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,3584,0.06278400288687812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,3584,0.02540622154871623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,3072,0.046705779102113515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,3072,0.0540231102042728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,3584,0.05398044321272108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,3072,0.055608888467152916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,3072,0.023123555713229712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,3584,0.0616133345497979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,2560,0.04654666781425476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,2048,0.03422933485772874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,2048,0.039473778671688504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,4096,0.06943022542529635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,2560,0.04173244370354546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,1536,0.027024888330035742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,2048,0.04132444328731961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,2560,0.02088088956144121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,1536,0.033879111210505165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,1024,0.01750577820671929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,4096,0.06934755378299289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,1024,0.023734221855799358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,768,0.013977777626779346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,1024,0.014193778236707052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,1024,0.0265600002474255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,2560,0.04855644371774462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,512,0.011672889192899069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,768,0.013284444808959961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,1536,0.016000888413853116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,768,0.022456000248591106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,256,0.008964444200197855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,512,0.012422222230169507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,512,0.019211555520693462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,2048,0.018398221996095445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,256,0.014193778236707052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,128,0.008011555506123437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,256,0.016197333733240765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,128,0.011750222080283694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,128,0.013461333182122974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,64,0.0069982219073507525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,768,0.019745777050654095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,512,0.016519111063745286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,32,0.013575110998418597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,256,0.01181422256761127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,8192,0.04918577935960558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,65536,0.7536746660868326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,128,0.014474666780895658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,64,0.013434666726324292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,65536,1.0042515860663521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,51200,0.6448648770650228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,65536,0.8599457740783691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,16384,0.21673866113026938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,51200,0.2536799907684326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,51200,0.782075564066569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,12288,0.13769600126478407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,16384,0.24550223350524902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,16384,0.25717245207892525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,12288,0.17155644628736708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,12288,0.15948000219133165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,10240,0.12712621688842773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,1536,0.031457778480317854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,10240,0.14459466934204102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,10240,0.054117331902186074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,8192,0.10417422321107651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,65536,0.3230862352583143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,8192,0.04247644543647766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,8192,0.10792355404959784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,7168,0.10251022047466701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,16384,0.08514310916264851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,51200,0.6454222467210557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,6144,0.07478844457202487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,7168,0.03844622108671401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,7168,0.09599910842047797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,6144,0.08972444136937459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,6144,0.03349688980314467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,5120,0.06075555748409695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,32,0.007287999822033777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,5120,0.07591466771231757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,5120,0.07270310984717475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,5120,0.029172443681293066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,12288,0.06410044431686401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,4096,0.06058222055435181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,7168,0.08896978033913507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,6144,0.085372441344791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,3584,0.04324266645643446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,4096,0.024541333317756653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,3072,0.03481422199143304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,3584,0.053973333703147046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,3584,0.022600889205932617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,10240,0.13389955626593694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,2560,0.029855112234751385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,3584,0.05483288897408379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,3072,0.047785778840382896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,3072,0.04845688740412394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,2048,0.02497600018978119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,2560,0.041778667105568774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,2560,0.04249155521392822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,4096,0.06093155675464206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,2048,0.017122666041056316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,1536,0.01958311100800832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,2048,0.036714666419559054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,4096,0.04853866828812493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,3072,0.020626667473051283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,1024,0.014358222484588623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,1536,0.014868444866604276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,8192,0.11610844400193955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,1024,0.021149333980348375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,768,0.012069332930776807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,1024,0.02335022224320306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,768,0.01792977750301361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,768,0.019874667127927143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,1536,0.028718223174413044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,512,0.009415110780133141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,1536,0.029896001021067303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,2560,0.01864266726705763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,512,0.011516444385051727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,512,0.014878221684032016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,1024,0.013140444126394061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,2048,0.035396443472968206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,128,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,256,0.013844444519943662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,128,0.012129777835475074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,768,0.01237066669596566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,64,0.006325333482689328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,128,0.013745778136783175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,256,0.007291555404663086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,64,0.012111999922328524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,512,0.01717688971095615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,32,0.012443555725945367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,256,0.011030221978823343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,65536,0.6552809079488119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,65536,0.293034659491645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,51200,0.500307559967041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,128,0.010967111421955956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,65536,0.8150551054212781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,16384,0.16777155134412977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,65536,0.8546089066399468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,51200,0.24839732382032606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,12288,0.1307342184914483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,16384,0.1948497825198703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,51200,0.6226497756110297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,16384,0.20371645026736787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,10240,0.10551377799775864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,12288,0.060658666822645396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,32,0.006611555400821898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,10240,0.051629334688186646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,8192,0.08717777993943955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,256,0.012787555654843649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,8192,0.10040266646279229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,12288,0.14672799905141196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,12288,0.15467822551727295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,10240,0.12404444482591416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,10240,0.13123555978139242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,7168,0.08825421995586819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,51200,0.6742151048448352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,6144,0.06265866756439209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,7168,0.035749332772360906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,7168,0.0943031112353007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,5120,0.05257155497868856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,6144,0.0313715570502811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,6144,0.07658044497172038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,8192,0.10673422283596462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,6144,0.08401599857542251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,5120,0.02753777801990509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,5120,0.06467733118269178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,5120,0.07078666819466485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,4096,0.023129777775870428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,16384,0.07840177747938368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,3584,0.047247999244266085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,3584,0.052271998590893216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,4096,0.04359022114011976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,3584,0.03842488924662272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,4096,0.05855200025770399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,3072,0.01958755486541324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,3072,0.04710222283999125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,3584,0.021363554729355708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,8192,0.039934221241209246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,2560,0.036107556687461004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,2560,0.01789955629242791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,2048,0.025254221426116094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,2560,0.04058311051792569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,2048,0.01607288916905721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,2048,0.030603554513719346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,1536,0.019339554839664035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,2048,0.0350488887892829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,2560,0.02903555499182807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,1024,0.014846222268210517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,1536,0.024545777175161574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,7168,0.07494399944941203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,1024,0.012832889126406776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,768,0.011648000114493899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,1024,0.02238666680124071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,768,0.01571911076704661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,768,0.01890222231547038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,512,0.009657777845859528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,1536,0.014488889111412896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,3072,0.03329599897066752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,512,0.013167111410035027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,256,0.007416889071464539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,512,0.016880000631014507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,4096,0.05317777726385328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,256,0.011338666909270816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,256,0.013479111095269522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,256,0.010507555471526252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,1536,0.028424001402325098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,128,0.0107351111041175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,3072,0.04183200001716614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,128,0.010595555934641095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,64,0.0107360002067354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,128,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,512,0.011038222246699862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,32,0.010779555473062726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,65536,0.4712382422553168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,64,0.007287110719415877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,65536,0.24595467249552408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,32,0.007872000336647034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,65536,0.6631982061598036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,65536,0.6433528794182671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,1024,0.018580444984965853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,51200,0.19617511166466606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,51200,0.5182933277553982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,51200,0.5025359789530436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,16384,0.1640124453438653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,12288,0.09822311004002889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,51200,0.3442097769843207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,768,0.012129777835475074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,12288,0.05155555407206217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,12288,0.1244444449742635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,12288,0.1271724436018202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,16384,0.12938488854302302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,8192,0.06416977776421441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,10240,0.04433955417739021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,10240,0.10802221960491604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,16384,0.06605155600441827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,8192,0.03380799955791897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,16384,0.16581866476270887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,8192,0.0836648874812656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,8192,0.0886311133702596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,6144,0.04836533466974894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,7168,0.030517333083682593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,7168,0.07421866390440199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,7168,0.07812711265352038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,5120,0.03926577832963731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,6144,0.06418844726350573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,128,0.01275466630856196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,6144,0.0692746639251709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,4096,0.029567999972237483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,5120,0.02369955513212416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,5120,0.053808887799580894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,7168,0.057342224650912814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,4096,0.02000977761215634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,4096,0.044999112685521446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,4096,0.04970133304595947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,3584,0.018556444181336295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,6144,0.026869333452648584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,3584,0.04490577843454149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,3072,0.03562666641341315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,5120,0.05898844533496433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,3072,0.0399635566605462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,3072,0.01715288890732659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,10240,0.08454666535059611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,2560,0.0309840010272132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,3584,0.039825777212778725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,2048,0.017246221502621967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,10240,0.10398933622572158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,2048,0.014471999473041959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,1536,0.013615111509958902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,2048,0.029323554701275293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,2560,0.0203128887547387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,1536,0.021475555168257818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,1024,0.010544000400437249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,2560,0.015850666496488783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,2560,0.034840001000298396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,1024,0.016114667057991028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,1024,0.011516444385051727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,768,0.008657777474986183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,2048,0.025938666529125635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,768,0.01074577785200543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,768,0.013625777430004544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,512,0.007116444408893585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,768,0.017269333203633625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,1536,0.012810666528013019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,1536,0.024728000164031982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,256,0.005961777849329843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,3072,0.022839110758569505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,512,0.01460533340771993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,128,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,1024,0.019850666324297588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,256,0.01330133361948861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,128,0.009385777844323052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,64,0.004967999955018361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,128,0.011513777905040316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,32,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,64,0.009377777576446533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,32,0.009353777600659264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,256,0.010045333041085137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,65536,0.3416915469699436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,3584,0.02663644486003452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,65536,0.19793689250946045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,65536,0.5166968769497341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,51200,0.25744088490804035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,65536,0.4997013409932454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,512,0.011745777395036487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,16384,0.09002844492594402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,51200,0.4077466593848334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,256,0.010055999788973067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,51200,0.39113865958319766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,16384,0.13372088803185359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,128,0.009704888694816166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,12288,0.06839111116197374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,16384,0.13049066066741943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,10240,0.05661333269543118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,12288,0.04234310984611511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,12288,0.10122044218911065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,512,0.010377777947319878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,12288,0.10051911407046848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,8192,0.04515377680460612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,10240,0.036879112323125206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,51200,0.15456444687313503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,10240,0.08559644222259521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,7168,0.03897422220971849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,8192,0.06787288851208158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,16384,0.053620444403754346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,8192,0.07095022333992852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,6144,0.03211999932924906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,7168,0.059283554553985596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,5120,0.028811554114023846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,6144,0.022264000442292955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,6144,0.05225511060820686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,6144,0.05533688929345873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,4096,0.022591999835438196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,5120,0.019557333654827543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,10240,0.08406577507654826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,5120,0.04774844315316942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,4096,0.016911110944218107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,4096,0.03642488850487603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,4096,0.0407608879937066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,7168,0.024440889557202656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,3584,0.032850666178597346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,3584,0.015822221835454304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,7168,0.06252800093756782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,3584,0.03654399845335219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,2560,0.015678221980730694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,3072,0.014534221755133735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,3072,0.0291608903143141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,5120,0.04422400063938565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,2048,0.013482666677898832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,2560,0.013470222552617392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,2560,0.025212445192866858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,3584,0.020251555575264823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,1536,0.011561777856614856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,2048,0.01241422196229299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,2048,0.02163644466135237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,2048,0.02425777746571435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,1024,0.009348444640636444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,1536,0.017991110682487488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,1536,0.0203457772731781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,1024,0.010374222364690568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,3072,0.03203199969397651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,768,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,1024,0.016538666354285348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,8192,0.027492443720499676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,512,0.006596444381607904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,768,0.010080888867378235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,768,0.014800000521871777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,3072,0.01805688937505086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,512,0.009696889254781935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,256,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,512,0.012851555314328937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,1024,0.013573333621025085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,256,0.008609777523411645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,128,0.004899555610285865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,256,0.01108000013563368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,2560,0.027980443504121568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,128,0.009010666774378883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,768,0.011807111402352652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,32,0.006275555739800136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,1536,0.01146488885084788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,512,0.010044444766309526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,32,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,65536,0.2643520037333171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,256,0.009078222016493479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,65536,0.19201689296298555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,64,0.004275555411974589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,128,0.007977777885066139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,128,0.010403555300500657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,65536,0.49539735582139754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,16384,0.06809866428375244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,64,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,51200,0.1530426608191596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,51200,0.3407635423872206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,12288,0.0517813331551022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,16384,0.05271822214126587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,51200,0.3894542323218451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,16384,0.1299146678712633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,12288,0.04168622361289131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,12288,0.09989421897464329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,10240,0.07164711422390409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,10240,0.03586755527390374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,8192,0.032638221979141235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,10240,0.08449421988593207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,8192,0.05804622173309326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,16384,0.11265244748857285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,8192,0.06993777884377374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,7168,0.03021777669588725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,10240,0.044212443961037524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,12288,0.08552355236477321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,7168,0.024225778049892847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,6144,0.025032000409232244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,7168,0.06271555688646105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,6144,0.044920888211992055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,5120,0.020678222179412842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,8192,0.026869333452648584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,51200,0.2127617730034722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,7168,0.051333334710862905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,4096,0.016881777180565726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,5120,0.019280888968043856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,5120,0.046762665112813316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,6144,0.021948445174429152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,4096,0.016885333591037326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,3584,0.015507555670208402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,4096,0.03951822386847602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,3072,0.013999111122555204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,3584,0.028632889191309612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,3584,0.035191112094455294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,65536,0.4423111014895969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,2560,0.012186666329701742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,3072,0.014496889379289416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,3072,0.031203554736243352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,2560,0.02186044388347202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,6144,0.054656889703538686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,2048,0.01037599974208408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,2560,0.02742133372359806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,2048,0.012313777373896705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,3584,0.015656888484954834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,1536,0.00903911143541336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,2048,0.023823999696307715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,3072,0.025202666719754536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,1536,0.011409777734014722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,1536,0.01998844411638048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,2560,0.013454222016864352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,4096,0.03172977765401205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,768,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,1024,0.01036888857682546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,1024,0.016199999385409884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,5120,0.0380968881977929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,768,0.010044444766309526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,768,0.014461333552996317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,1024,0.007305777735180325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,512,0.009167999857001835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,256,0.004660444541109933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,512,0.009532444179058075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,1024,0.012085333466529846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,256,0.009062222308582729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,128,0.00433688900536961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,2048,0.018972444865438674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,512,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,256,0.01075111081202825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,64,0.0038782221575578055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,128,0.00902844468752543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,128,0.010413332945770687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,512,0.013111111190583972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,32,0.0070017774899800616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,65536,0.18447289201948378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,768,0.010461333725187512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,65536,0.38442399766710067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,65536,0.14620710743798151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,32,0.004226666771703296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,65536,0.34678488307529026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,51200,0.14783377117580837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,64,0.00701333334048589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,51200,0.11754843923780654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,51200,0.3021048969692654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,51200,0.2720888985527886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,16384,0.10064178042941624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,256,0.007725333174069722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,16384,0.09335733122295803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,128,0.007084444165229797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,12288,0.033033778270085655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,10240,0.032297776805029974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,12288,0.07652088668611315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,12288,0.07341244485643175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,10240,0.06153244442409939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,8192,0.02504355543189579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,16384,0.04124088750945197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,8192,0.05211822191874186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,8192,0.019998222589492798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,7168,0.02156266735659705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,16384,0.049039999643961586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,7168,0.04550933175616794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,6144,0.019528888993793063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,10240,0.0294773346847958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,12288,0.03805955582194858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,6144,0.04010666741265191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,5120,0.01684888866212633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,1536,0.015615110596021017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,7168,0.0456008878019121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,10240,0.06388622522354126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,5120,0.014848889576064216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,4096,0.028282665544086035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,6144,0.016749334004190233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,8192,0.05200799968507555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,5120,0.03433955709139506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,4096,0.013171555267439948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,3584,0.012568000290128918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,4096,0.014111111561457316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,3584,0.012427555190192329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,3584,0.02552088929547204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,3584,0.026372444298532274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,6144,0.039953778187433876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,3072,0.02291733357641432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,3072,0.01183377785815133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,2560,0.010117332968446944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,4096,0.029139555162853662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,5120,0.0347377790345086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,7168,0.01849688920709822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,2560,0.011085333095656501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,2560,0.020114666885799833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,3072,0.011533333195580376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,2048,0.01723911033736335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,2560,0.02147377696302202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,2048,0.018533334136009216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,2048,0.010684444672531553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,1024,0.007095110913117726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,1536,0.01424088908566369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,1024,0.009409777820110321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,2048,0.008771555291281806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,768,0.00591466658645206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,1024,0.013356444736321768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,1024,0.01109333336353302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,512,0.005180444568395615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,768,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,768,0.009711999860074785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,768,0.011791999969217511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,256,0.004262222184075249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,512,0.008452444440788692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,1536,0.009992889232105678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,3072,0.02402488887310028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,128,0.004003555410438114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,256,0.008007110820876228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,256,0.009701333112186855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,1536,0.01614488826857673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,128,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,64,0.003644444462325838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,128,0.00889066689544254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,32,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,1536,0.007583111524581909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,65536,0.11020888884862264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,64,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,32,0.006592000110281839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,512,0.010888000329335531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,51200,0.09184800253974067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,65536,0.14442043834262425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,256,0.00703288863102595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,65536,0.3139457702636719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,16384,0.03658755620320638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,65536,0.3314302232530382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,51200,0.243248012330797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,128,0.008120888637171851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,51200,0.26034667756822377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,16384,0.040864888164732195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,16384,0.08952977922227646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,10240,0.02051911089155409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,512,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,12288,0.032891554964913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,12288,0.06878666745291816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,12288,0.06173955731921726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,8192,0.01688888834582435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,10240,0.05208711160553826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,10240,0.05871466795603434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,51200,0.11590045028262669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,7168,0.0151573336786694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,8192,0.042239109675089516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,8192,0.048756443791919284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,16384,0.08142489194869995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,7168,0.018165333403481376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,7168,0.03756355577045017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,7168,0.043191999197006226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,5120,0.011771555576059552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,6144,0.03301422132386102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,6144,0.03834844297832913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,10240,0.02849600050184462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,5120,0.02791555391417609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,5120,0.014998222390810648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,5120,0.03234577841228909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,8192,0.019837333096398246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,3584,0.009656888743241629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,12288,0.025568000144428674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,4096,0.013232889274756113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,4096,0.023294222023752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,3072,0.00832088871134652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,3584,0.012457778056462606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,3584,0.02129688858985901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,6144,0.016639111770523917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,2560,0.00797155582242542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,3072,0.011787555283970304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,3072,0.0225564440091451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,4096,0.010517333116796283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,2048,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,2560,0.011076444553004371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,2560,0.02020088831583659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,6144,0.01352355546421475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,4096,0.027553778555658128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,1536,0.005976888868543837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,3584,0.025103110406133864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,3072,0.0189057770702574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,1536,0.012147555748621622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,1024,0.004938666605287128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,1536,0.009943111075295342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,1536,0.01498311095767551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,2560,0.016740444633695815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,768,0.004621333132187526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,1024,0.009359999663299983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,1024,0.009724444813198512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,1024,0.012426666915416718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,2048,0.014516444669829475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,768,0.00868088917599784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,768,0.011071999867757162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,512,0.007524444825119442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,512,0.008341333104504479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,256,0.0041697778635554844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,256,0.006262222097979651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,256,0.0090515555606948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,128,0.0038808890514903595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,512,0.0042133331298828125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,768,0.009127111070685917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,64,0.0036275556517971885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,128,0.008057778080304464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,32,0.00360000009338061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,128,0.008676444490750631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,64,0.005647111270162795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,32,0.005650666852792104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,256,0.008174222376611497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,65536,0.2689342233869764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,51200,0.06850133339564006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,65536,0.14560533894432917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,65536,0.32815734545389813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,128,0.005674666828579373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,512,0.01037333326207267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,16384,0.02828888760672675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,51200,0.11593600114186604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,65536,0.08336799674563938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,12288,0.018572444717089336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,16384,0.041725334193971425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,16384,0.07194133599599202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,16384,0.0884551140997145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,12288,0.03327466713057624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,12288,0.06814489099714491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,10240,0.04621866676542494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,10240,0.058041777875688344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,8192,0.013362666798962487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,51200,0.25770311885409886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,8192,0.03692177931467692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,8192,0.04826400015089247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,7168,0.01204177737236023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,8192,0.021294222937689886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,7168,0.032953778902689614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,51200,0.2107724481158786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,10240,0.02860355708334181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,7168,0.019091554813914828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,2048,0.010564444793595208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,6144,0.02857511242230733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,5120,0.00941066692272822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,5120,0.024735111329290602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,6144,0.010760000182522668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,5120,0.015554666519165039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,5120,0.03223289052645365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,2048,0.01759111053413815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,6144,0.016844444804721408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,4096,0.020745777421527438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,4096,0.01347644461525811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,4096,0.02713866697417365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,7168,0.04277688927120633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,3584,0.0247662216424942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,3072,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,4096,0.0081386665503184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,6144,0.037308444579442344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,3072,0.016911110944218107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,2560,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,3072,0.012092444631788464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,3072,0.022525333695941504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,2560,0.011733333269755045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,2048,0.005579555614127054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,12288,0.054758220911026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,2048,0.01331999980741077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,2048,0.010796444283591377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,1536,0.005004444470008214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,2048,0.017216889394654166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,2560,0.015152888165579902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,1536,0.011114666859308878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,1024,0.004265777766704559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,1536,0.014872888724009195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,3584,0.01294133315483729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,1024,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,768,0.003912888881233004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,1024,0.012431999875439538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,768,0.008045333127180735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,768,0.01107911103301578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,3584,0.018935999936527677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,512,0.003535110917356279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,1536,0.010066666536860997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,10240,0.016159999701711867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,512,0.010047111246320937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,256,0.003504000190231535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,256,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,2560,0.019871999820073444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,768,0.00904977818330129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,128,0.0031315556002987754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,256,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,256,0.008838222258620793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,512,0.006983111302057902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,64,0.0029226665695508323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,128,0.008000888758235509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,1024,0.009404444032245213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,64,0.005285333428117964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,32,0.0053395554423332214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,128,0.005325333525737126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,65536,0.2691182295481364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,51200,0.062344888846079506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,16384,0.02626666592227088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,12288,0.01422133379512363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,65536,0.0772071083386739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,10240,0.013453333742088742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,51200,0.21019555462731254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,8192,0.012796444197495779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,12288,0.05371110969119602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,7168,0.011425777441925473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,10240,0.04523644513554043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,6144,0.010199999643696679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,8192,0.036710222562154136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,5120,0.00906044410334693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,7168,0.03257511059443156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,4096,0.007899555895063613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,6144,0.028304000695546467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,3584,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,5120,0.024648000796635944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,3072,0.006679111056857639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,16384,0.07042488786909315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,2560,0.0059262220230367445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,3584,0.01865688959757487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,2048,0.0053182223604785065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,3072,0.016933333542611863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,1536,0.004952888935804367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,512,0.008446222378147973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,2048,0.01313333296113544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,1024,0.004330666528807746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,768,0.003998222036494149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,1536,0.011238222320874533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,1024,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,512,0.0036595554815398324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,4096,0.020761777957280476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,256,0.0032808888289663526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,128,0.003002666764789157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,768,0.008017777568764156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,512,0.0070053330726093715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,256,0.005623111294375525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,32,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,128,0.0052915554907586836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,64,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,65536,0.07583644655015734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,51200,0.061624891228146024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,65536,0.2683795558081733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,16384,0.025067556235525344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,64,0.002895111011134254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,12288,0.01242844429281023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,2560,0.014885332849290637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,32,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,10240,0.011051555474599203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,8192,0.009724444813198512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,7168,0.009747555686367882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,12288,0.05350755651791891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,10240,0.04604977700445387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,8192,0.03661777906947666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,5120,0.01129155523247189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,6144,0.028397334946526423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,4096,0.009681777821646797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,5120,0.024544888072543677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,3584,0.008707555631796518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,16384,0.070269332991706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,3072,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,32,0.0053191110491752625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,2560,0.007320889168315464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,3072,0.016872000363137987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,7168,0.03199199835459391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,2048,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,1536,0.004594666676388847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,2560,0.015184889237085978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,1024,0.0042462220622433555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,2048,0.01309866706530253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,768,0.003867555409669876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,1536,0.01108000013563368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,512,0.0035502223504914176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,1024,0.00906755526860555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,4096,0.020626667473051283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,3584,0.018787556224399142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,6144,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,512,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,64,0.0026862221873468826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,256,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,32,0.0028417776856157514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,51200,0.21009422673119438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,64,0.0052915554907586836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,32,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,256,0.0031759999692440033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,768,0.008243555823961893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,128,0.0029093333416514923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,128,0.0053048887186580235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,16384,1.4453706741333008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,16384,4.438296000162761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,16384,4.808302985297309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,12288,3.3544125027126737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,12288,1.0686719682481554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,12288,3.4973333146837025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,51200,5.459915584988064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,10240,0.8680364290873209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,10240,3.0269376966688366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,8192,0.6860124270121256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,10240,2.837058597140842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,8192,2.181184980604384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,7168,0.6346373558044434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,8192,2.3686737484402127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,7168,2.0668267144097223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,7168,1.8925777011447484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,6144,0.5365991062588161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,6144,1.7371457417805989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,6144,1.6150986353556316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,5120,1.359231948852539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,51200,13.061783684624565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,4096,1.1291981803046334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,4096,0.3744577831692166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,5120,1.3169049157036674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,5120,0.45039643181694883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,3584,0.33814488516913516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,3584,1.055697758992513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,3584,0.9500462214152018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,3072,0.8718906508551704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,3072,0.7801244523790148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,4096,1.0485235850016277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,2560,0.29525865448845756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,2048,0.5118115743001302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,2048,0.22199377748701307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,3072,0.29570399390326607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,2048,0.5390657848782009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,1536,0.176801774236891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,1024,0.2898720105489095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,1024,0.1433768934673733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,1024,0.3193262153201633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,768,0.22760266727871367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,1536,0.4140346580081516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,768,0.1253475613064236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,512,0.18301333321465385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,1536,0.43857065836588544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,128,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,2560,0.6665502124362522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,512,0.11044533385170831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,512,0.2179351117875841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,256,0.09685955444971721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,128,0.09535377555423313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,128,0.18485955397288004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,2560,0.8150924576653374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,32,0.07971466912163629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,64,0.1885946724149916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,128,0.09092177947362264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,51200,19.729337904188366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,32,0.19173244635264078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,768,0.25358133845859104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,256,0.1343404452006022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,256,0.19593332873450386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,64,0.07614933119879828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,16384,3.287088818020291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,16384,3.703325483534071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,16384,1.1005075242784288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,51200,4.189891391330295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,65536,5.932226816813151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,12288,2.465033849080404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,12288,0.7939395374721951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,12288,2.821443557739258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,10240,0.6833004421657987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,10240,2.311026679144965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,51200,10.163095262315538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,10240,2.148934258355035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,8192,0.5408408906724718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,8192,1.6375706990559895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,8192,1.7977981567382812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,7168,0.48351287841796875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,7168,1.540056864420573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,6144,0.42020267910427517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,6144,1.365565299987793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,6144,1.2426764170328777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,5120,0.36202043957180446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,5120,1.163040902879503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,7168,1.434905793931749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,4096,0.8208364380730523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,4096,0.3035919931199816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,3584,0.7894044452243381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,3584,0.27277421951293945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,5120,1.0372195773654513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,3072,0.7554480234781901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,3072,0.23765688472323945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,3072,0.6251475546095107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,2560,0.5632061958312988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,51200,15.929239908854166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,65536,12.88006591796875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,2560,0.21451021565331352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,2048,0.4554462432861328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,4096,0.8248035642835828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,2048,0.17691644032796225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,1536,0.4117128849029541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,2048,0.4240453243255615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,3584,0.7240284283955892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,1024,0.24713688426547578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,2560,0.5238240030076768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,1024,0.11081777678595649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,1024,0.2514266702863905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,768,0.18265067206488717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,768,0.09722399711608887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,512,0.13124799728393555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,768,0.19980977641211617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,256,0.09437777598698933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,512,0.0878408882353041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,128,0.0733226670159234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,256,0.07801777786678739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,256,0.15455910894605848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,64,0.06246844265196058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,128,0.07679644558164808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,32,0.06491021977530585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,128,0.1458266708585951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,65536,19.532141791449654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,64,0.14871999952528211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,32,0.1509502198961046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,1536,0.14281688796149358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,512,0.17149599393208823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,16384,1.0532746844821506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,1536,0.33588798840840656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,65536,1.748753759596083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,16384,0.358420451482137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,16384,1.204646216498481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,51200,1.3664639790852864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,51200,3.4180177052815757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,12288,0.8712906837463379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,12288,0.2708195580376519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,65536,4.338350084092882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,10240,0.8029297722710503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,8192,0.5439537896050347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,10240,0.7086977958679199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,12288,1.0026817321777344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,8192,0.22605154249403211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,7168,0.5341191291809082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,8192,0.5713360044691298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,7168,0.16507111655341253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,6144,0.48094489839341903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,5120,0.37005776829189724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,6144,0.49469953113132054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,5120,0.13232799371083578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,4096,0.25341065724690753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,5120,0.4073066711425781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,4096,0.1039866672621833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,4096,0.2810453308953179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,3584,0.09170133537716335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,3584,0.24413598908318412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,3072,0.20872622066073945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,3072,0.08134488926993476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,3072,0.20792088243696424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,3584,0.24764890140957305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,2560,0.19682489501105416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,65536,7.127384609646267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,2048,0.14140000608232287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,2560,0.17618932988908556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,2048,0.06434755855136447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,2048,0.14464799563090006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,1536,0.050126221444871694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,1536,0.11531289418538411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,10240,0.2295937803056505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,2560,0.07297777467303805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,1024,0.08773333496517605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,768,0.06801778078079224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,1536,0.0992586678928799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,768,0.03854044609599643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,512,0.04442310995525784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,768,0.07027110788557264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,256,0.03386666708522373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,512,0.035403556293911405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,512,0.0601102246178521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,256,0.031698667340808444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,7168,0.47335465749104816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,256,0.05380355649524265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,64,0.024887111451890733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,1024,0.04204977883232964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,128,0.05093422200944689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,64,0.051743110020955406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,32,0.05255022313859728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,1024,0.07932621902889676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,128,0.02887466549873352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,51200,5.236453162299262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,128,0.031324442889955305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,32,0.02602044410175747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,65536,1.2180328369140625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,16384,0.8499946594238281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,51200,0.976943122016059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,16384,0.2819226582845052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,51200,2.4775252872043185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,65536,3.1791324615478516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,6144,0.14893599351247153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,16384,0.9389449225531684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,12288,0.21864267190297446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,12288,0.6542231241861979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,10240,0.17521688673231336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,12288,0.6685804261101617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,51200,3.778917524549696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,8192,0.36974843343098956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,8192,0.14345422055986193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,7168,0.3508524364895291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,8192,0.4126266638437907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,65536,4.742494371202257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,10240,0.43509689966837567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,7168,0.12790132893456355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,7168,0.37422932518853086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,6144,0.11086399687661065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,10240,0.5389608807033962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,5120,0.09794044494628906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,4096,0.20031378004286024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,5120,0.2590000099605984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,4096,0.08123822344674005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,3584,0.07128266493479411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,3072,0.16191555394066706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,3584,0.18897067175971138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,3072,0.0647324456108941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,3072,0.1586915519502428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,6144,0.30871555540296763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,2560,0.05607999695671929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,2560,0.1357040007909139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,2048,0.11620888445112441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,3584,0.18043466409047446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,1536,0.07980977826648288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,2048,0.04892444279458788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,4096,0.20871910783979628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,1024,0.05799555778503418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,1536,0.04069955481423272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,1536,0.0890995528962877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,1024,0.03379911184310913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,6144,0.288719998465644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,1024,0.06775289111667208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,5120,0.22165511714087593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,768,0.031369778845045306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,256,0.026177777184380427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,512,0.028840889533360798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,512,0.046816001335779824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,128,0.020991999242040846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,256,0.02610844373703003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,256,0.04207377632459005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,64,0.017492444978819955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,128,0.02539377742343479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,32,0.01816444430086348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,768,0.051002668009863965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,2560,0.13019111421373156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,64,0.04034488730960422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,32,0.04077511032422384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,768,0.054746667544047035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,2048,0.11215110619862874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,128,0.039697776238123574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,512,0.04016888803905911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,65536,1.0758790969848633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,16384,0.6572675704956055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,51200,0.7943378024631076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,51200,2.0156035953097873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,16384,0.7357208993699816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,65536,2.6588728162977433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,12288,0.17989867263370088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,10240,0.37508267826504177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,12288,0.6376826498243544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,10240,0.15450578265719944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,8192,0.30864977836608887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,8192,0.12464000119103326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,7168,0.2608915434943305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,65536,4.006114535861545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,7168,0.1111111111111111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,7168,0.30804088380601674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,6144,0.09763199753231472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,6144,0.2675120035807292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,5120,0.23418577512105307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,51200,3.4398167928059897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,4096,0.18794755140940347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,5120,0.2192417780558268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,6144,0.2775173452165392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,4096,0.0700204438633389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,3584,0.15200532807244196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,16384,0.30331288443671334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,3584,0.061031997203826904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,3584,0.15465244981977674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,5120,0.08354044622845119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,3072,0.13581600454118517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,2560,0.12101866139305963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,2560,0.048806223604414195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,2560,0.1156960063510471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,2048,0.09135378069347805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,3072,0.14049955209096274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,2048,0.04231111208597819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,1536,0.0737395551469591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,2048,0.09546399778789944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,1024,0.045687112543318004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,1536,0.07591822412278917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,4096,0.17587021986643472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,10240,0.4515368673536513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,768,0.03837688763936361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,1024,0.028910222980711196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,12288,0.5360595385233561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,512,0.02805244591501024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,768,0.04683644572893778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,1536,0.03516799873775906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,256,0.02106933295726776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,512,0.024146666129430134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,512,0.03900444507598877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,256,0.021885333789719477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,256,0.03451822201410929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,128,0.02137511140770382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,128,0.03228444523281521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,64,0.01589866644806332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,768,0.026591999663247004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,32,0.015903110305468243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,3072,0.05606222152709961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,64,0.032568001084857516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,32,0.03257066673702664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,128,0.018567999203999836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,1024,0.05762755870819092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,8192,0.349132431877984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,16384,0.5112257533603245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,51200,0.6511084238688151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,16384,0.1946586635377672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,16384,0.5983049074808756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,65536,2.187730577256944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,51200,1.828205320570204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,12288,0.42056621445549863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,12288,0.44213154580858016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,65536,0.8552097744411893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,10240,0.13116533226437038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,10240,0.4889769024319119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,51200,2.5299352010091147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,8192,0.10035733381907146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,8192,0.2825591034359402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,7168,0.27583644125196666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,65536,3.421056959364149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,7168,0.08967110845777725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,7168,0.26343199941847056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,12288,0.14766044086880153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,5120,0.2073031134075589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,10240,0.3786337905459934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,6144,0.25939644707573783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,4096,0.147871110174391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,8192,0.2777004506852892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,3584,0.12066755029890273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,4096,0.14401955074734157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,3584,0.051500442955229014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,3072,0.10407021972868179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,3584,0.1276951101091173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,6144,0.23115287886725533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,3072,0.0458551115459866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,6144,0.08635466628604466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,3072,0.11260178354051377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,2560,0.04052711195415921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,5120,0.06776977909935845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,2048,0.03510844376352098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,4096,0.0570933355225457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,1536,0.028792887926101685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,1536,0.06269866890377469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,2560,0.08973333570692275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,1024,0.023863111933072407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,1024,0.04718755682309469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,768,0.0317191117339664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,2560,0.0960364474190606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,1536,0.05635822481579251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,5120,0.1827902264065213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,768,0.03840355409516229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,256,0.018240888913472492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,1024,0.03939111034075419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,256,0.027793778313530818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,128,0.015104000767072042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,2048,0.07967999908659193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,128,0.018043556147151522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,512,0.02402488887310028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,2048,0.07895288864771526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,64,0.012867555850081973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,32,0.012801777985360889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,512,0.019659555620617337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,64,0.02641955514748891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,32,0.026058667235904272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,256,0.018187556001875136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,768,0.02160266704029507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,65536,0.8525333404541016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,512,0.03177244464556376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,51200,1.5670319663153753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,65536,2.1056151919894748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,51200,0.684117317199707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,16384,0.19472800360785592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,12288,0.4183315436045329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,12288,0.14128622743818495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,12288,0.423891544342041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,10240,0.11806844340430366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,10240,0.31919911172654897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,8192,0.2439840104844835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,16384,0.5271689097086588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,8192,0.0936666660838657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,8192,0.25165510177612305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,65536,2.7968097262912326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,7168,0.08233333296246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,7168,0.22116977638668486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,6144,0.16890133751763237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,128,0.02644266684850057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,6144,0.07224177651935153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,10240,0.3127768834431966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,5120,0.1613182226816813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,5120,0.06194755766126844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,4096,0.11754755179087321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,4096,0.05171644356515673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,5120,0.15386845005883112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,3584,0.10767110851075913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,51200,2.377643585205078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,3584,0.11348977353837754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,16384,0.5254142019483778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,2560,0.08282666736178927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,3072,0.04150399896833632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,3072,0.09910400046242608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,2560,0.03692888882425096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,6144,0.190628449122111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,2560,0.0862026678191291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,1536,0.04559200008710226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,2048,0.030969778696695965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,7168,0.21863467163509795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,1024,0.03534400131967332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,3072,0.08777955505583022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,1536,0.05697421895133125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,1024,0.02156977852185567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,1024,0.042621334393819175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,4096,0.12914133071899414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,768,0.03478577733039856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,3584,0.0462719996770223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,512,0.028519110547171697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,1536,0.026127111580636766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,512,0.01776088939772712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,768,0.0258933338854048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,256,0.016583111551072862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,256,0.02481599979930454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,2048,0.062881780995263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,128,0.016521778371598985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,128,0.023759111762046814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,32,0.012857777376969656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,2048,0.07085866398281522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,64,0.023758222659428913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,256,0.015831111205948722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,32,0.023677332533730402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,128,0.013809777796268463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,768,0.019328888919618394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,512,0.020647999313142564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,65536,0.7282266616821289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,16384,0.4667813513014052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,51200,1.2910666995578342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,51200,0.5701582166883681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,16384,0.1579564412434896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,16384,0.6160311169094509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,64,0.01236177815331353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,12288,0.11767200628916423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,12288,0.33417778544955784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,51200,1.9122159745958116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,8192,0.2173759937286377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,10240,0.2759599950578478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,65536,1.7175661722819011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,7168,0.19269866413540312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,8192,0.21847821606530082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,12288,0.3449697759416368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,6144,0.15959289338853624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,7168,0.19262844986385771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,10240,0.27673421965705025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,5120,0.11996532811058892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,10240,0.09990044434865315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,6144,0.16822666592068145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,4096,0.10496444172329372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,5120,0.05527199970351326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,5120,0.14071999655829534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,8192,0.0828346676296658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,4096,0.04606755574544271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,4096,0.11321066485510932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,3072,0.07662044631110297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,3584,0.04111466805140177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,7168,0.07344266441133288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,3072,0.036954667833116316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,3072,0.08730933401319717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,2560,0.03284444411595663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,2560,0.07496266894870333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,2048,0.027830223242441814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,3584,0.08713689115312363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,2048,0.06201510959201389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,1536,0.02369511127471924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,2560,0.07076088587443034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,1024,0.028099556763966877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,1536,0.050195554892222084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,6144,0.06422666708628337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,1024,0.02016177773475647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,2048,0.05574755536185371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,512,0.017625777257813346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,768,0.01827822294500139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,768,0.031091554297341242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,512,0.016572443975342643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,1536,0.04084800018204583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,512,0.025223111112912495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,256,0.015426667200194465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,128,0.012005333271291522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,768,0.02186311119132572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,65536,2.4892569647894964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,128,0.015624889069133334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,256,0.013797333670987023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,32,0.011047999891969891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,3584,0.09988977511723836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,256,0.02200711104604933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,32,0.020969778299331665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,1024,0.037689778539869524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,64,0.01033866653839747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,64,0.0210853334930208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,65536,0.5900275442335341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,51200,0.967940436469184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,65536,1.5484346813625758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,128,0.021020443903075323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,12288,0.25841066572401256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,16384,0.3945724434322781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,12288,0.2844115628136529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,10240,0.19470132721794975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,51200,1.5698702070448134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,10240,0.08869777785407172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,8192,0.18055466810862222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,10240,0.23770043585035536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,16384,0.16267377800411648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,7168,0.12683467070261636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,8192,0.18887110551198324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,12288,0.10185511244667901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,7168,0.062407109472486705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,7168,0.16615288787417942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,6144,0.054732445213529796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,6144,0.14280088742574057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,16384,0.3662044472164578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,5120,0.048698667022917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,51200,0.46467288335164386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,4096,0.03922755519549052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,6144,0.11413066916995579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,4096,0.09694044457541572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,5120,0.10836178064346313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,3072,0.0680933329794142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,8192,0.07435822486877441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,3072,0.07536533143785265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,2560,0.058317330148484975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,3072,0.03171199891302321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,3584,0.07347022162543403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,2560,0.028231110837724473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,2560,0.0647288892004225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,3584,0.08622488710615371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,2048,0.024559110403060913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,2048,0.05372088816430834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,5120,0.12133244673411052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,1024,0.023386667172114056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,65536,2.12296273973253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,1024,0.01811644434928894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,768,0.019996444384257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,1536,0.03837066557672288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,768,0.027151111099455092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,768,0.016187555260128446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,4096,0.09351466761695014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,1536,0.04338933361901177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,512,0.015058666467666626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,256,0.011732444167137146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,512,0.02235200007756551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,256,0.013873777455753751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,2048,0.04626311196221245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,256,0.019310222731696237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,3584,0.035459554857677884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,64,0.009366222553782994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,128,0.01419911119672987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,32,0.009681777821646797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,128,0.018583999739752877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,64,0.01829866733815935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,1536,0.021339555581410725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,512,0.014986667368147107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,1024,0.032688889238569475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,65536,1.0857715606689453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,51200,0.8959759606255425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,16384,0.28448088963826496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,51200,0.3802151150173611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,32,0.01826133330663045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,16384,0.11804889308081733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,16384,0.41419288847181535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,51200,1.2982791264851887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,128,0.010728889041476779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,12288,0.23165154457092285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,8192,0.13164622253841823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,10240,0.20881066057417128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,8192,0.060489780373043485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,8192,0.15567378203074136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,12288,0.21123822530110678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,7168,0.1368231111102634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,6144,0.11502844757503933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,65536,0.48131646050347227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,10240,0.18023733297983804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,5120,0.10260711113611858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,10240,0.07555199993981256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,6144,0.12012088298797607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,5120,0.04072533382309808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,4096,0.07152266634835137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,65536,1.7679182688395183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,7168,0.05357244279649523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,12288,0.08821511268615723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,7168,0.13737244076199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,3584,0.061583108372158475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,3072,0.051262223058276706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,3584,0.030638221237394545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,3584,0.07171199719111125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,3072,0.0276826669772466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,5120,0.10062755478752984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,2560,0.024640889631377325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,2560,0.05446399913893806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,4096,0.03360711203681098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,4096,0.08132177591323853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,2048,0.02144444485505422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,6144,0.04761866728464762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,2048,0.04569688770506117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,1024,0.021055110626750525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,1536,0.018568888306617737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,2560,0.04143911268975999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,1536,0.036878221564822726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,768,0.016650666793187458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,1024,0.015573332707087198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,512,0.013155555559529198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,768,0.014504000544548035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,256,0.010066666536860997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,2048,0.03895911243226793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,512,0.013456888496875763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,512,0.01903733279969957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,256,0.012796444197495779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,1536,0.029814223448435467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,256,0.01642311116059621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,128,0.012757333616415659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,32,0.007795555724038019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,3072,0.06328088707394071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,64,0.01586666703224182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,1024,0.027433777848879497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,32,0.015648888217078317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,768,0.02310488952530755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,128,0.008608889248636033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,64,0.00777599960565567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,65536,0.4290026558770074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,51200,0.727275530497233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,51200,0.3459644317626953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,16384,0.10650400320688884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,12288,0.19239555464850533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,16384,0.31254132588704425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,65536,1.4041546715630426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,51200,1.0898319880167644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,10240,0.17344978120591906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,12288,0.2089395523071289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,8192,0.11662399768829346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,10240,0.18103378348880342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,128,0.015862221519152325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,7168,0.11779022216796875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,16384,0.2554311222500271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,6144,0.10699555608961318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,7168,0.1242106623119778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,10240,0.07167288992140028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,6144,0.10793333583407932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,5120,0.08749510844548543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,8192,0.054874665207333036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,5120,0.036102222071753606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,4096,0.06255111429426405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,5120,0.09120266967349583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,7168,0.04837422238455879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,4096,0.030553778012593586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,4096,0.07343910800086127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,3584,0.02791733211941189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,3584,0.06529688835144043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,6144,0.042338665988710195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,3072,0.025243555506070454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,2560,0.04219555523660448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,3072,0.05777511331770155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,2048,0.03362933463520474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,8192,0.14060533046722412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,2560,0.04948888884650337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,3584,0.055656890074412026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,1536,0.026543999711672466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,2048,0.042111999458736844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,1024,0.016734222571055096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,3072,0.046909332275390625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,1536,0.0336808893415663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,768,0.014518222875065274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,1024,0.014998222390810648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,65536,0.8744728830125598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,768,0.013987556099891663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,768,0.02106133268939124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,2560,0.022952889402707417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,256,0.009011555876996782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,512,0.012849777936935425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,512,0.01776444415251414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,128,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,256,0.012120889292822944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,1536,0.01753777762254079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,12288,0.08722755644056533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,64,0.007276444799370236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,128,0.012081777883900536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,128,0.014471999473041959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,1024,0.025241777300834656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,64,0.014501333236694336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,512,0.011686222420798408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,32,0.014525334040323893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,2048,0.0199297782447603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,65536,0.3775955571068658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,256,0.015150222513410779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,51200,0.7200897534688314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,32,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,51200,0.30896798769632977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,16384,0.09029599693086411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,12288,0.1660577721065945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,16384,0.27370667457580566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,12288,0.0741457806693183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,12288,0.18617244561513266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,65536,0.8077537748548719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,10240,0.06008266740375095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,8192,0.09475911325878567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,16384,0.23222043779161242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,7168,0.10016710890663995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,65536,1.1853378083970811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,8192,0.12640533182356092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,51200,0.8838666809929742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,7168,0.11012177997165257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,10240,0.13917156060536703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,6144,0.03689510954750909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,6144,0.09701333443323772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,4096,0.0546248886320326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,10240,0.15902666250864664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,5120,0.032214221027162336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,8192,0.04871288935343424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,4096,0.027664000789324444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,4096,0.06589511368009779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,3072,0.0424124432934655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,3584,0.05834311246871948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,5120,0.07013244761361016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,2560,0.035169776942994856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,5120,0.0815804468260871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,3584,0.05283555719587538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,2560,0.020298666424221463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,3584,0.025056888659795124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,2048,0.03770133190684848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,1536,0.02275466587808397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,3072,0.0513626668188307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,1536,0.01591822173860338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,1536,0.03012088934580485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,2560,0.044575111733542554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,2048,0.02827377782927619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,768,0.012747555143303342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,1024,0.013796444568369122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,7168,0.04192444350984362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,512,0.01013777818944719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,768,0.013109332985348172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,2048,0.017888888716697693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,768,0.018803555104467604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,256,0.008008889026112026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,3072,0.02271644439962175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,512,0.015791111522250705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,256,0.011622221933470832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,256,0.013791999883121915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,128,0.011395555403497485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,128,0.013081777426931592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,32,0.006640888750553131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,1024,0.022683555881182354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,6144,0.09160444471571182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,32,0.013407111167907715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,512,0.012104888757069906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,128,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,65536,0.7010657522413465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,1024,0.015488889482286243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,64,0.0063795554969045855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,16384,0.17471733358171251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,51200,0.28167910046047634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,64,0.01312622262371911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,16384,0.21436799897087944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,51200,0.7266800138685438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,12288,0.062336001131269664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,12288,0.16279378202226427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,65536,0.31979110505845815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,10240,0.052686221069759794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,8192,0.0888035562303331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,51200,0.5283430947197808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,7168,0.07212266657087538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,8192,0.04101955559518602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,16384,0.07821689049402873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,12288,0.13284177250332302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,6144,0.06904977560043335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,7168,0.09678488969802856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,10240,0.11385156048668756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,6144,0.032461331950293645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,5120,0.05242400036917793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,6144,0.0840853320227729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,5120,0.028370665179358587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,5120,0.07104621993170844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,4096,0.024433778391944036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,3584,0.040687110688951277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,4096,0.05708533525466919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,65536,0.9577848646375867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,3072,0.03670577870474921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,8192,0.11024177736706203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,3584,0.05127733283572727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,3072,0.02016266683737437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,2560,0.029986666308508977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,3072,0.04537689023547702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,2048,0.024875554773542616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,2560,0.018593778212865193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,2560,0.03937155670589871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,1536,0.02066044509410858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,2048,0.03342666559749179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,1536,0.014541332920392355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,3584,0.02236000034544203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,1024,0.01568000018596649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,7168,0.036166220903396606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,768,0.012598222328556908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,1024,0.012792000340090858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,1024,0.020041777027977836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,512,0.010538666612572141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,768,0.016897777716318767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,256,0.00832444429397583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,512,0.011191111471917896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,512,0.014370666609870063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,128,0.00647199981742435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,256,0.010760889285140567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,1536,0.02683555583159129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,128,0.010749333434634738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,128,0.011800000237094032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,32,0.006595555692911148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,4096,0.04910844564437866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,32,0.011854222251309289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,65536,0.562903086344401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,256,0.012337778177526263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,64,0.005967999911970562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,51200,0.4018426736195882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,64,0.011499555574523078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,65536,0.7602604230244955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,16384,0.1504711045159234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,51200,0.23639289538065592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,51200,0.60043912463718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,16384,0.07328089078267415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,16384,0.1816257768207126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,10240,0.09507911072836982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,12288,0.13959733645121256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,65536,0.31169867515563965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,768,0.012099555797047086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,10240,0.11571999390920003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,2048,0.01628444426589542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,8192,0.03889333208401998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,8192,0.09376266929838394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,12288,0.10780888795852661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,6144,0.06057600180308024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,7168,0.03474666674931844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,12288,0.057809776730007596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,5120,0.04795022143257988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,6144,0.030730668041441176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,6144,0.07236355543136597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,10240,0.05075200067626106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,5120,0.06018222040600247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,10240,0.13655911551581487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,3584,0.03274844421280755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,4096,0.022699554761250813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,3584,0.021008888880411785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,3584,0.044212443961037524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,3072,0.03920533259709676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,3072,0.01937333411640591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,7168,0.06483288606007893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,2560,0.02436711059676276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,4096,0.049245334333843656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,2048,0.019354666272799175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,2560,0.017834666702482436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,2560,0.03398222062322829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,1536,0.01607111096382141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,2048,0.02866311205757989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,4096,0.03796799977620443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,1536,0.02348444362481435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,1024,0.017659554878870647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,1024,0.012269333004951477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,768,0.009946666657924652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,5120,0.02683199942111969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,1024,0.011735111474990845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,768,0.011600000162919363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,512,0.01292622254954444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,3072,0.026745777991082933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,512,0.010832889212502373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,8192,0.07828622394137912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,256,0.011005333728260465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,256,0.010399999717871347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,128,0.005631111148330901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,128,0.010377777947319878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,128,0.01039644413524204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,7168,0.08256177769766913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,32,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,256,0.006311111152172089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,64,0.01034577770365609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,32,0.010403555300500657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,65536,0.43707021077473956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,512,0.007924444145626491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,65536,0.2369706630706787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,51200,0.33778932359483504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,64,0.005208888815508948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,2048,0.015839111473825242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,16384,0.11226844787597656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,51200,0.4808150927225749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,1536,0.01418133411142561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,16384,0.06602311134338379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,16384,0.1527128881878323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,10240,0.07842489083607991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,12288,0.11541599697536892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,10240,0.04354844490687052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,65536,0.6041039890713161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,8192,0.06233688857820299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,51200,0.18814399507310656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,7168,0.0513608886135949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,768,0.01495911180973053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,12288,0.09145244624879624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,8192,0.07762666543324788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,6144,0.04475288920932346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,7168,0.02940622303220961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,12288,0.04948800139957004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,7168,0.0689200030432807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,4096,0.029108444849650066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,5120,0.0229413327243593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,5120,0.050423999627431236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,3584,0.027083555857340496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,8192,0.03264177838961283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,3584,0.03740888833999634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,3584,0.01814311080508762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,3072,0.02309422194957733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,5120,0.04055644406212701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,6144,0.02641511129008399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,3072,0.016760889026853774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,3072,0.03305066625277201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,2048,0.017250667015711468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,10240,0.09681777821646796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,4096,0.019660444723235238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,4096,0.04170133339034187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,1536,0.013487111363146039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,2560,0.02846133377816942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,2048,0.024267555938826665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,1536,0.012471111284361945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,1536,0.02012711101108127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,768,0.008676444490750631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,1024,0.015164444843928019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,2560,0.020594666401545208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,512,0.00721244431204266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,6144,0.059644447432623975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,768,0.012947555217478009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,256,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,512,0.010045333041085137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,512,0.011076444553004371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,2560,0.015501333607567681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,2048,0.014357333381970724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,1024,0.010427555276287926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,256,0.009700444837411245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,64,0.004597333156400257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,32,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,128,0.00903911143541336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,1024,0.011347555451922946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,64,0.0090515555606948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,65536,0.3024551073710124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,65536,0.1873697837193807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,51200,0.22273688846164277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,65536,0.46848535537719727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,128,0.005320888840489917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,16384,0.08054400152630277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,51200,0.15286400583055285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,128,0.009565333525339762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,12288,0.05719377597173055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,51200,0.3676915433671739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,16384,0.12053155899047852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,10240,0.04767022199100918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,12288,0.09173778030607437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,32,0.00906844437122345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,8192,0.038718223571777344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,10240,0.03570933474434747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,10240,0.07650488615036011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,7168,0.03225422236654494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,8192,0.02646488944689433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,6144,0.029233776860766943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,7168,0.02426311042573717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,7168,0.054472890165117055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,6144,0.021532444490326777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,16384,0.05152622196409437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,6144,0.04734844300482008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,4096,0.019852444529533386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,12288,0.04103199972046746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,5120,0.0403457780679067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,3584,0.017644445101420086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,4096,0.03347200155258179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,256,0.009703999592198266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,3584,0.0303457776705424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,3072,0.015942222542232938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,5120,0.02390577726893955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,3072,0.014482667048772177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,3072,0.026540444956885442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,5120,0.018858666221300762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,2048,0.011852444046073489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,768,0.01071466671095954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,2560,0.02342044479317135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,2048,0.012083555261294046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,1536,0.0099653336736891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,8192,0.06143199735217624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,1536,0.01110311100880305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,1536,0.016656888855828177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,2560,0.013516444298956128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,1024,0.010042666561073726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,2560,0.013283555706342062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,1024,0.0127688886390792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,4096,0.01657777859105004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,512,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,768,0.009690666364298927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,2048,0.01997155613369412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,768,0.01108266661564509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,256,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,512,0.009907555249002244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,128,0.0042275554604000514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,1024,0.008001777860853408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,256,0.008112000094519721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,64,0.003920000046491623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,128,0.00872444444232517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,128,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,768,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,64,0.007632888853549957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,65536,0.24169333775838217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,32,0.007629333270920648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,3584,0.015504888362354703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,51200,0.19767378436194527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,65536,0.18533155653211805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,256,0.008774222599135505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,51200,0.1481706698735555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,51200,0.3257173432244195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,32,0.004261333495378494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,12288,0.04047733214166429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,12288,0.08187555604510836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,10240,0.041913777589797974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,512,0.009300444689061906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,8192,0.03195377853181627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,10240,0.06872622172037761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,65536,0.41485778490702313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,7168,0.027884443600972492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,8192,0.055452442831463285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,16384,0.10866489013036092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,6144,0.025840888420740765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,7168,0.02347466680738661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,7168,0.04916977882385254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,5120,0.021087111698256597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,6144,0.021239999267790053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,6144,0.043268445465299815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,5120,0.018914666440751817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,16384,0.06614666514926486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,5120,0.03689510954750909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,3584,0.016314667132165697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,4096,0.016529776983790927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,8192,0.02631644407908122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,4096,0.030448890394634668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,3072,0.014410666293568082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,3584,0.015368888775507608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,2560,0.012645333177513547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,3072,0.014154665999942355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,3072,0.0244159996509552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,2048,0.010662222074137794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,2560,0.021359110871950786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,4096,0.017418666018380057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,1536,0.009015999734401703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,2048,0.018453333112928603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,12288,0.05143288771311442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,1024,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,3584,0.027445332871543035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,1024,0.011785777906576792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,2560,0.01313866674900055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,10240,0.03578222129080031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,2048,0.012091555529170565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,768,0.010386666489972008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,16384,0.0507279998726315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,1536,0.015173332558737861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,256,0.004575110971927643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,512,0.009179555707507664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,256,0.00868444475862715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,1024,0.010070222119490305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,128,0.004224888980388641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,128,0.008671999805503422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,64,0.00388355553150177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,128,0.007027555671003129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,512,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,1536,0.011064888702498542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,64,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,32,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,65536,0.17541066805521646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,51200,0.14150577121310765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,65536,0.14007643858591715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,65536,0.3368364440070258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,51200,0.11263999674055312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,16384,0.04497244291835361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,768,0.009735999835862054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,16384,0.03964089022742377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,12288,0.035079999102486506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,16384,0.09024800194634332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,512,0.008987555901209513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,10240,0.02773244513405694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,12288,0.06722844309277005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,256,0.007615110940403408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,10240,0.05635644329918755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,32,0.0042088888585567474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,8192,0.04593066705597771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,7168,0.019592000378502738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,8192,0.01995822290579478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,12288,0.03143289022975498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,7168,0.017952000101407368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,6144,0.0169840008020401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,7168,0.04090311129887899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,768,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,6144,0.016507556041081745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,10240,0.027172444595230952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,6144,0.036173333724339805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,4096,0.01293955577744378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,5120,0.014818666709793938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,3584,0.011622221933470832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,4096,0.013055110971132914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,4096,0.025460445218616064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,3584,0.012164444559150271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,3584,0.022877333892716303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,3072,0.011757333245542316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,3072,0.020649777518378366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,2560,0.009360888765917884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,8192,0.021340444684028625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,2560,0.011119999819331698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,2560,0.01790488925245073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,5120,0.03021866745418972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,1536,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,2048,0.015519110692871941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,1024,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,3072,0.010763555765151978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,1536,0.01277511070171992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,768,0.004892444445027245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,1024,0.010098666780524785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,5120,0.015509333875444202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,768,0.00868977771864997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,512,0.004571555389298333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,768,0.00924711094962226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,256,0.0037671112351947357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,512,0.008341333104504479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,2048,0.010430222584141625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,256,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,256,0.00666844430896971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,1536,0.009703999592198266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,64,0.003536888708670934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,128,0.007979555262459649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,1024,0.00906044410334693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,128,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,65536,0.10558933019638062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,2048,0.008325333396593729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,32,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,51200,0.08506577544742161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,512,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,128,0.003572444534964032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,16384,0.034536888202031456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,51200,0.1113884449005127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,51200,0.2306417889065213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,32,0.003527111063400904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,16384,0.03907644417550828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,16384,0.07705777883529663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,64,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,12288,0.030204445123672485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,12288,0.05933066871431139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,10240,0.026846221751636926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,51200,0.26575732231140137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,65536,0.29372800721062553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,8192,0.019440889358520508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,8192,0.04009066687689887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,12288,0.022641777992248535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,6144,0.011935999823941125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,7168,0.03592977921168009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,10240,0.01830311119556427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,6144,0.016224000189039443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,5120,0.010057777994208867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,6144,0.031846221950319074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,4096,0.008718222379684448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,5120,0.027098667290475633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,7168,0.013268444273206921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,3584,0.00833422193924586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,4096,0.012883555557992725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,3584,0.012435555458068848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,3584,0.020282665888468426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,3072,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,8192,0.014509333504570855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,10240,0.049309333165486656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,3072,0.011531555818186866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,3072,0.018258665998776753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,2560,0.01110311100880305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,2560,0.016199999385409884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,7168,0.01755466726091173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,1536,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,2048,0.014013333453072442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,1536,0.009689778089523315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,1024,0.004611555486917496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,65536,0.13987910747528076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,1024,0.009000889129108852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,768,0.004269333349333869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,4096,0.022464889619085524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,2048,0.010411555568377176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,768,0.008441777692900764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,768,0.00868266655339135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,1536,0.011776888536082374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,5120,0.01484444406297472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,512,0.008170666793982187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,1024,0.009355555805895064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,256,0.005994666781690385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,128,0.0032222221295038858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,512,0.0038808890514903595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,128,0.005590222362014983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,64,0.0028799999919202593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,128,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,2560,0.006681777950790193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,512,0.007288888924651676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,65536,0.07948178052902222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,64,0.005628444254398346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,32,0.005633777628342311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,256,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,51200,0.06271377536985609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,65536,0.13818577925364176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,16384,0.02611200014750163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,65536,0.25336888101365834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,51200,0.11118843820359971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,12288,0.016544888416926067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,16384,0.03997066617012024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,32,0.0031999999450312722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,12288,0.0317191117339664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,12288,0.05063733458518982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,10240,0.028572443458769057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,10240,0.04315733247333103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,8192,0.0196142229768965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,8192,0.034669333034091525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,7168,0.011786667009194693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,51200,0.198052446047465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,6144,0.01075111081202825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,7168,0.030971556901931763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,16384,0.06648000081380208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,5120,0.009358222285906473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,6144,0.017263111140992906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,8192,0.013152889079517789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,4096,0.00795733349190818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,5120,0.02342044479317135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,4096,0.013256889250543384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,4096,0.01995377739270528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,3584,0.007288888924651676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,7168,0.01807733376820882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,3072,0.006614222294754452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,3584,0.012440889245933957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,10240,0.014564444621404013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,2560,0.006239111224810283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,3072,0.011791999969217511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,6144,0.02734222180313534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,5120,0.01519733336236742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,2560,0.01109777804878023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,2560,0.014496000276671516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,1536,0.004903111192915174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,2048,0.010395555860466426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,2048,0.012708444562223224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,1536,0.009715555442704095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,1536,0.01074399964676963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,256,0.003536000019974179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,1024,0.009022222624884712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,2048,0.005975111077229182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,3072,0.01629599928855896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,512,0.0035217776894569397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,768,0.00868977771864997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,768,0.007768889268239339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,256,0.0032097777972618737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,512,0.008278222547637092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,512,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,128,0.0028728888266616394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,1024,0.004230222354332606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,256,0.005626666463083691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,64,0.002879111096262932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,128,0.007931555310885111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,3584,0.018283555905024212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,32,0.002846222163902389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,768,0.0038782221575578055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,65536,0.07183911403020223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,64,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,51200,0.058351112736595996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,2048,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,32,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,16384,0.022858666049109563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,12288,0.014314666390419006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,10240,0.012825777961148156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,65536,0.2537422180175781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,51200,0.1978133387035794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,256,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,12288,0.051515556044048734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,7168,0.010563555690977307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,10240,0.04288977715704176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,8192,0.03448444604873657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,7168,0.03080000148879157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,6144,0.027152889304690894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,4096,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,5120,0.023218666513760883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,3584,0.0069982219073507525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,3584,0.01793333391348521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,8192,0.011161777708265515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,4096,0.019765333996878732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,3072,0.006648889018429651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,2560,0.00600355573826366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,6144,0.010710222025712332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,2048,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,128,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,1536,0.004580444345871608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,2560,0.014552889598740472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,1024,0.0042035554846127825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,2048,0.012711111042234631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,768,0.0038773334688610504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,1536,0.01072355525361167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,512,0.003532444437344869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,1024,0.008685333033402761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,16384,0.06634577777650622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,768,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,128,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,512,0.0069751110341813825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,64,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,256,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,1024,0.008691555923885768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,32,0.002883555574549569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,128,0.0052728888889153796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,65536,0.07648444175720215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,64,0.005294222384691238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,51200,0.058435552650027804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,16384,0.022871999277008906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,65536,0.2525600062476264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,256,0.003245333416594399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,12288,0.013165333204799227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,51200,0.20298311445448133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,10240,0.011738667057620155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,16384,0.06634222136603461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,8192,0.013225778109497495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,7168,0.009682666924264696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,6144,0.009015111459626092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,5120,0.009567110902733272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,5120,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,7168,0.03086844417783949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,4096,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,32,0.004943111290534337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,3584,0.008725333544943068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,5120,0.023573334018389385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,3072,0.00832622249921163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,4096,0.019719999697473314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,2560,0.006945778098371293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,3584,0.017940445078743827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,2048,0.005312888986534542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,3072,0.016234666109085083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,12288,0.05069777700636122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,2560,0.014472888575659858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,8192,0.03460977805985345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,1536,0.010827555424637265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,768,0.0035866668654812705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,768,0.00774133370982276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,512,0.0032391111469931076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,3072,0.016275554895401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,256,0.0032684444967243406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,2048,0.012549333274364471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,128,0.0029848888516426086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,6144,0.026709333062171936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,64,0.0028702221396896574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,1024,0.008672888908121321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,32,0.0028453332682450614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,64,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,32,0.004959110998445087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,1536,0.004577777865860197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,10240,0.0429075558980306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,1024,0.003933333274390963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,512,0.006950221955776215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,256,0.00563822231358952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,128,0.005241777747869492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,16384,1.3618835873074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,16384,4.168651580810547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,16384,4.63874774509006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,12288,3.274919933742947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,51200,5.31608623928494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,16384,4.127274831136067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,12288,1.0126320521036785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,12288,3.400408850775825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,10240,0.8257777955796984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,10240,2.6804462009006076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,12288,2.9982079399956594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,10240,3.218923568725586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,8192,0.6753964424133301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,10240,2.4967208438449435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,8192,2.188025792439779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,8192,2.067030164930556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,51200,12.826634724934896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,8192,2.049611621432834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,7168,0.5875377655029297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,7168,2.362460454305013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,7168,1.868944803873698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,6144,0.5068453152974447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,6144,1.8008783128526475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,7168,1.7474764717949762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,6144,1.5280240376790364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,5120,1.3153190612792969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,5120,0.4299573368496365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,6144,1.5319564607408311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,4096,0.9963066313001844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,5120,1.3100639979044597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,4096,0.3688062296973334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,4096,1.0332924524943035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,3584,0.9528444078233508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,5120,1.2190453211466472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,3584,0.331679105758667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,3584,0.8978648715549045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,3072,0.8669484456380209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,3584,0.8740391201443143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,4096,0.9320479498969184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,3072,0.3465030988057454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,2560,0.598340458340115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,2560,0.2543040116628011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,3072,0.7318933275010852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,2560,0.6485902468363444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,2048,0.4745484458075629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,2560,0.6233919991387261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,51200,14.009146796332466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,2048,0.21018132898542616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,51200,20.00730726453993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,2048,0.5631377961900499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,1536,0.1708728869756063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,1536,0.42040620909796816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,1024,0.13471021917131212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,1024,0.29067643483479816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,768,0.243980434205797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,1536,0.426163567437066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,2048,0.5074053340488011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,768,0.25051289134555393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,768,0.2445333268907335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,1024,0.31765156322055393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,512,0.2134782208336724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,512,0.10597066746817695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,512,0.19421777460310194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,3072,0.7954995367262098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,256,0.0927964448928833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,512,0.1692239973280165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,128,0.08873599767684937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,256,0.15386754936642116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,1536,0.4086462126837836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,64,0.074617776605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,128,0.18135289351145426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,128,0.1177884472741021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,32,0.08451288938522339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,1024,0.3001920117272271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,32,0.18865066104465059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,768,0.12018844816419814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,128,0.09102755784988403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,256,0.10746577713224624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,64,0.18551377455393472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,256,0.19285066922505698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,65536,5.2676311069064665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,51200,4.0199767218695746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,16384,3.722632090250651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,51200,10.299803839789496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,16384,3.9588445027669272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,16384,1.0166026221381295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,65536,13.234378390842013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,12288,2.5456915961371527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,16384,3.2509822845458984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,12288,2.6632934146457248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,12288,0.7647671169704862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,12288,2.3356552124023438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,51200,11.682967291937935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,51200,15.098312377929688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,10240,0.648721800910102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,10240,2.1007181803385415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,8192,1.9852701822916667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,8192,0.5229333241780599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,10240,2.3602256774902344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,8192,1.6275030771891277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,7168,0.46045955022176105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,65536,19.98164537217882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,8192,1.568348460727268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,10240,1.9902879926893446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,7168,1.4401689105563695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,65536,15.22664048936632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,6144,0.4192986753251817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,7168,1.3570684856838648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,6144,1.4046693378024633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,6144,1.2257395850287545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,5120,1.0409937964545357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,7168,1.430175993177626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,6144,1.1636986202663846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,4096,0.8537973297966851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,4096,0.2868186632792155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,5120,0.9749902089436849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,4096,0.7987857924567329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,3584,0.6897875467936198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,5120,0.3494711187150743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,3584,0.26136710908677846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,3584,0.7084337870279948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,3072,0.5425910949707031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,5120,1.0429484049479167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,3072,0.23255554835001627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,3584,0.6453573438856337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,2560,0.4867386817932129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,3072,0.6293306880527073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,2560,0.20482310983869764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,3072,0.5788008901807997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,2560,0.5294684304131402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,4096,0.754982206556532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,2560,0.4753431214226617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,2048,0.1725848913192749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,2048,0.4206639925638835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,1536,0.137898670302497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,1024,0.191302220026652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,1536,0.33126666810777455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,1536,0.3113208876715766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,1024,0.10731466611226399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,768,0.1709831158320109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,2048,0.3936346636878119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,1536,0.27773600154452854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,768,0.1945777734120687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,512,0.1368995507558187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,2048,0.4056559933556451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,512,0.08484355608622234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,256,0.0962373349401686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,512,0.15333955817752415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,1024,0.24306045638190377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,128,0.07043288813696967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,256,0.0748035576608446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,256,0.12203200658162434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,128,0.14308888382381862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,128,0.07303555806477864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,128,0.09388177924686009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,32,0.0612693362765842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,512,0.1690995560752021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,64,0.14641333950890437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,768,0.09381244579950969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,32,0.14805066585540771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,64,0.05875555674235026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,1024,0.24869865841335723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,768,0.19792711734771729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,65536,1.5799101723564997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,256,0.15205244223276773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,16384,1.1520409054226346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,51200,1.2226044336954753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,51200,3.3485245174831815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,16384,0.33614222208658856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,65536,4.2560780843098955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,16384,1.2011973063151042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,12288,0.9458817376030816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,51200,5.3354064093695746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,12288,0.25858932071261936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,51200,3.508347617255317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,65536,4.515574137369792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,12288,0.8426639768812391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,12288,0.717525323232015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,10240,0.22442756758795845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,16384,1.0038941701253254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,10240,0.648881753285726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,10240,0.6907475789388021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,8192,0.1763777732849121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,10240,0.580633799235026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,7168,0.45225154028998477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,8192,0.46638933817545575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,7168,0.464979542626275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,8192,0.5362897978888618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,6144,0.38300177786085343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,6144,0.406384891933865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,65536,6.426775190565322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,5120,0.3415866692860921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,5120,0.12604888280232748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,5120,0.33245600594414604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,5120,0.2962951130337185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,4096,0.2717706627315945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,4096,0.10515199767218696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,4096,0.26417244805230033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,8192,0.5483653280470107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,3584,0.23089599609375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,7168,0.15813599692450628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,3584,0.08933777941597833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,3584,0.23236799240112305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,3584,0.2221004433102078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,7168,0.40789421399434406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,3072,0.08084444204966228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,3072,0.20456177658504912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,6144,0.13838577270507812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,3072,0.18764266702863905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,4096,0.24191999435424805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,2560,0.17356533474392363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,2560,0.1635999944474962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,2048,0.05968888600667318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,2048,0.14280800024668375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,1536,0.10369688934750027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,2560,0.16606667306688097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,6144,0.36232354905870223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,1536,0.04936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,1024,0.08090222544140287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,1024,0.0871279968155755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,3072,0.19553777906629774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,768,0.059133330980936684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,1536,0.11346756087409125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,768,0.06914755370881823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,1536,0.10845688978830974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,512,0.04596711198488871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,768,0.06921866867277357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,1024,0.0825386643409729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,512,0.059141331248813205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,1024,0.04099644554985894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,256,0.031417777140935264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,512,0.034444444709353976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,512,0.05700000127156576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,256,0.053103999959097974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,256,0.030623111459943984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,128,0.026770666241645813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,128,0.05013511247105069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,2048,0.13600888517167833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,64,0.022287999590237934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,64,0.05104710989528232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,256,0.04514666729503208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,32,0.02259822189807892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,32,0.05204177896181742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,128,0.03671111000908746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,128,0.030231999026404485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,2048,0.14093511634402806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,768,0.03745511174201965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,65536,1.1764631271362305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,2560,0.0682577755716112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,16384,0.8171093198988172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,65536,3.3555164337158203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,51200,2.522213406032986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,16384,0.2550720108879937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,16384,0.9610498216417102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,16384,0.750026649898953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,12288,0.5627724329630533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,51200,0.9453137715657552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,51200,2.607170740763346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,65536,4.716551038953993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,65536,3.3388328552246094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,12288,0.2055377827750312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,10240,0.5522906515333388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,10240,0.18434133794572619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,12288,0.5412906540764703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,8192,0.3981466558244493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,10240,0.5670853190951878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,10240,0.44762754440307617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,8192,0.13837511009640163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,51200,3.6704533894856772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,8192,0.437349345948961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,8192,0.3513946798112657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,6144,0.30106043815612793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,7168,0.31727645132276744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,12288,0.6352879736158583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,6144,0.10525155729717678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,6144,0.30604354540507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,5120,0.23936178949144152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,6144,0.2681368986765544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,4096,0.17863110701243082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,5120,0.2569039927588569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,5120,0.22625954945882162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,7168,0.1250844399134318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,4096,0.07616621918148465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,7168,0.35814311769273544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,4096,0.19012710783216688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,3584,0.06844177511003283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,3584,0.1787982251909044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,3584,0.1643706692589654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,5120,0.09351733658048843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,3072,0.06045689185460409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,7168,0.3698879877726237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,2560,0.12628711594475642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,3072,0.14787556065453425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,4096,0.20229955514272055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,2048,0.11119643847147624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,2560,0.05438755618201362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,2560,0.1342995564142863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,2560,0.12603644529978433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,3072,0.14951554934183756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,2048,0.11081155803468491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,2048,0.10476977957619561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,1536,0.03984444340070089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,1536,0.08802222543292576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,3072,0.15670400195651585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,1536,0.08545244402355617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,768,0.04263200031386482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,1024,0.03296355406443278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,3584,0.16470577981736925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,1536,0.0781502193874783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,512,0.034656888908810086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,768,0.05428355601098803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,1024,0.06091199980841743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,512,0.04632711079385546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,512,0.02823822365866767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,1024,0.06793155272801717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,1024,0.06657155354817708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,256,0.04131644301944309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,128,0.022357333037588332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,768,0.03051111102104187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,256,0.036045332749684654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,128,0.025286222497622173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,64,0.018941332896550495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,512,0.04528977804713779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,32,0.01975733372900221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,64,0.03957155678007338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,32,0.040058665805392794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,256,0.025241777300834656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,128,0.039086222648620605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,128,0.030050665140151978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,256,0.024288000331984624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,2048,0.04683644572893778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,65536,0.9829351637098523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,65536,2.7995067172580295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,768,0.05403377612431844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,51200,2.4581449296739364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,16384,0.6190569135877821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,16384,0.2203235493765937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,51200,0.7624293433295356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,16384,0.759153789944119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,16384,0.6439733505249023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,51200,3.3410542805989585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,12288,0.4909955660502116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,12288,0.17148799366421172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,12288,0.541368908352322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,12288,0.45597420798407656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,10240,0.14190844694773355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,51200,2.2639973958333335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,8192,0.3595546616448297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,10240,0.3858959939744737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,65536,2.9540721045600047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,8192,0.12058044804467095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,8192,0.342137787077162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,8192,0.3112986617618137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,10240,0.48474401897854275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,6144,0.26483111911349827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,65536,4.1445494757758246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,10240,0.44516531626383465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,6144,0.09081511365042792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,5120,0.22604356871710884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,7168,0.29158933957417804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,5120,0.21646755271487764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,5120,0.07964799801508586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,5120,0.20992710855272081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,4096,0.17030844423505995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,7168,0.2731422318352593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,3584,0.15425599945916071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,4096,0.17349689536624482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,6144,0.2573191059960259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,4096,0.1617342233657837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,3072,0.12628266546461317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,6144,0.23740887641906738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,3584,0.14367822806040445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,7168,0.10288444492552017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,3072,0.05305244525273641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,3072,0.1268977721532186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,4096,0.06655289067162408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,2560,0.11413689454396565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,2560,0.046281778150134616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,2048,0.08718933661778767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,2560,0.10921955770916408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,2048,0.04053066505326165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,7168,0.2997857729593913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,1536,0.06824977530373467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,3072,0.1324977742301093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,1536,0.034396443102094866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,1536,0.07510133584340413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,1024,0.052410665485594005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,3584,0.05972533093558418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,1024,0.056806219948662647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,1024,0.028186665640936956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,768,0.03464711043569777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,1024,0.05800444549984402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,2048,0.09111111031638251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,2560,0.10598488648732503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,512,0.02955289019478692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,768,0.025693333811230127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,768,0.045736889044443764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,256,0.026956443985303242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,512,0.023388443721665278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,512,0.03860088851716783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,512,0.03980266716745164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,128,0.01885777711868286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,256,0.02103288968404134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,256,0.033913777934180364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,256,0.03186933199564616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,64,0.015814221567577787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,128,0.02125688890616099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,128,0.031910220781962075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,32,0.015581332974963717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,128,0.02642755541536543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,64,0.03195822238922119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,32,0.032058666149775185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,768,0.048453334305021495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,1536,0.07513333029217191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,3584,0.15333955817752415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,2048,0.09439022011227077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,65536,0.7955448892381457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,65536,2.073915481567383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,16384,0.6387146843804253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,16384,0.5979208946228027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,51200,2.6338693830702042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,51200,0.6250773535834419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,65536,3.387583202785916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,65536,2.348175048828125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,12288,0.42956267462836367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,12288,0.15551643901401097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,12288,0.4231128957536485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,10240,0.3626382350921631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,12288,0.3915431234571669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,16384,0.5289928648206923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,16384,0.18753955099317762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,8192,0.2777644528283013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,8192,0.27814221382141113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,8192,0.2530844476487902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,51200,1.5877822240193684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,7168,0.24134222666422525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,10240,0.35588354534573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,7168,0.24658134248521593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,7168,0.22325423028733996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,6144,0.2108231120639377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,8192,0.09873333242204453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,51200,1.8294302622477214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,6144,0.07718222008811103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,6144,0.1908951070573595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,5120,0.17812621593475342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,5120,0.06485866838031344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,5120,0.16329422261979845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,4096,0.12983555263943142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,4096,0.05451911025577121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,4096,0.1419644488228692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,10240,0.3526311185624864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,4096,0.13433599472045898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,6144,0.2123804489771525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,3072,0.09153778023189968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,3584,0.049830221467547946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,3584,0.12561955716874865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,3584,0.119576891263326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,7168,0.0888142254617479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,3072,0.04430222180154589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,3072,0.10998221900728013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,3072,0.10581244362725152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,2560,0.03884622123506334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,2560,0.09436177545123631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,10240,0.11814578374226888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,5120,0.16732354958852133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,1536,0.05384177631802029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,2048,0.03457244568400913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,2048,0.07662755250930786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,1536,0.027694221999910142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,1024,0.03720533185535007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,2048,0.07113511032528348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,3584,0.11252089341481526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,1024,0.046712001164754234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,768,0.030766222212049697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,1024,0.023077333966890972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,2048,0.07783910963270399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,1024,0.048489779233932495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,512,0.02125599980354309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,768,0.02091377807988061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,768,0.038034667571385704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,1536,0.0621982216835022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,768,0.040714667903052434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,256,0.01718222267097897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,512,0.019522666931152344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,512,0.03401244348949856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,256,0.01791111131509145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,2560,0.09151555432213677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,256,0.026879111925760906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,128,0.01867999964290195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,128,0.026293334033754136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,64,0.01275644451379776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,128,0.022830223043759663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,32,0.012778666284349231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,2560,0.08800088697009617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,1536,0.0643084446589152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,512,0.031065778599845037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,32,0.025736888249715168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,128,0.014819555812411837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,65536,0.7014302147759331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,64,0.02618666655487484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,256,0.027552000350422327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,65536,2.2735733456081815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,51200,0.568061351776123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,16384,0.5081520080566406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,51200,2.133836534288194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,65536,2.7405147552490234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,16384,0.5411093499925401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,51200,1.51690493689643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,16384,0.4655493100484212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,12288,0.1228133307562934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,12288,0.3915662235683865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,10240,0.315536896387736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,12288,0.32285600238376194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,10240,0.1281413369708591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,8192,0.2526133325364855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,10240,0.31488532490200466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,65536,2.033063040839301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,12288,0.3756817711724176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,7168,0.22293689515855578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,8192,0.2644960085550944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,8192,0.22923199335734049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,51200,1.5748817655775282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,7168,0.07787022325727674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,6144,0.17370488908555773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,7168,0.21873511208428276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,6144,0.06704622507095337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,6144,0.18698755900065103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,6144,0.17028621832529703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,8192,0.08455110920800103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,5120,0.05758577585220337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,4096,0.12247822019788955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,5120,0.14013155301411948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,4096,0.12660888830820718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,10240,0.2696719964345296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,4096,0.11780977249145508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,3584,0.10237866640090942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,5120,0.14390488465627035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,3584,0.04358844293488396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,16384,0.15926755799187556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,3072,0.09834844536251491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,4096,0.047909332646263965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,3072,0.03877244393030802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,3072,0.09285066525141399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,3584,0.11230488618214925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,7168,0.1940551069047716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,2560,0.03462400039037069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,2560,0.08454666535059611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,2560,0.08003999789555867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,5120,0.1578577756881714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,2048,0.029734220769670274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,2048,0.06821866830190022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,2560,0.07034666670693292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,1536,0.055607113573286265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,1024,0.02972266740269131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,3584,0.10660444365607367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,2048,0.06111022498872545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,1024,0.02130311065249973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,1024,0.04350577791531881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,1536,0.04404711061053806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,768,0.03453777896033393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,1536,0.025211556090248957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,768,0.036398222049077354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,512,0.018394667241308425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,512,0.028251555230882432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,1024,0.04216622312863668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,256,0.014482667048772177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,512,0.01757333344883389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,2048,0.06952089071273804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,256,0.01605333387851715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,256,0.024432000186708238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,128,0.012079111403889127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,256,0.02392622166209751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,768,0.019183110859658983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,1536,0.05571555429034763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,64,0.01124888906876246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,128,0.02344088918632931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,128,0.020581333173645865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,64,0.023341332872708637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,512,0.02993511160214742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,32,0.02304711110062069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,3072,0.09417421950234307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,128,0.016348444753222995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,32,0.011381333072980246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,65536,0.6910693380567762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,768,0.02555822167131636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,16384,0.4381413459777832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,51200,0.5437324311998155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,51200,1.42376889122857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,65536,1.8285181257459853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,16384,0.5108088917202419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,16384,0.42236799663967556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,51200,1.3204853269788954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,51200,2.0265911950005426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,12288,0.11530400647057427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,10240,0.2950177722507053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,12288,0.2935813268025716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,65536,1.5943821801079645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,16384,0.18542667229970297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,10240,0.2525786558787028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,8192,0.2447128825717502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,12288,0.31978845596313477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,8192,0.2157440053092109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,8192,0.19628355238172743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,7168,0.19023288620842826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,7168,0.17321599854363334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,7168,0.07105244530571832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,6144,0.17118399673038057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,7168,0.19898666275872123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,6144,0.16387377844916448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,6144,0.06056444512473213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,5120,0.14178755548265246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,65536,2.662469228108724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,5120,0.13889066378275552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,5120,0.13121599621242946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,4096,0.1023377776145935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,8192,0.0779786639743381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,4096,0.11077600055270725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,4096,0.04363110992643568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,6144,0.15098133352067736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,3584,0.09298044443130493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,10240,0.2765715652041965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,3584,0.039098666773902044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,3072,0.07659111420313518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,5120,0.05261244376500448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,3584,0.09501155879762437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,2560,0.06945244471232097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,3072,0.034846223062939115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,3072,0.0859759979777866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,12288,0.32615823215908474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,2560,0.03159288896454705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,4096,0.10636978016959296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,2048,0.05474755499098036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,10240,0.0980328917503357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,3584,0.09871821933322483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,2048,0.06131466892030504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,1536,0.04701688885688782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,2048,0.06283022297753228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,1536,0.022683555881182354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,3072,0.0846453309059143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,1024,0.030424889591005113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,1536,0.05039377676116096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,2560,0.07397333118650648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,768,0.02334755493534936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,1024,0.037201778756247625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,1024,0.04009777638647292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,2048,0.02663199934694502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,768,0.017662222186724346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,512,0.017844445175594754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,1536,0.04904177784919739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,512,0.025076443950335186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,512,0.027388445205158655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,512,0.016176000237464905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,1024,0.018960000740157235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,128,0.012097777591811286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,256,0.014832889040311178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,256,0.021703999903466966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,256,0.02185955478085412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,768,0.030424889591005113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,128,0.02092355489730835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,32,0.010405333505736457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,128,0.01881866653760274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,64,0.020978666014141504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,256,0.014172444740931192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,2560,0.07264977693557739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,64,0.010029333333174387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,768,0.03357777661747403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,65536,0.5814462237887913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,65536,1.6227555804782445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,32,0.02053511142730713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,16384,0.3139217694600423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,51200,0.46542220645480686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,65536,1.5444435543484156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,16384,0.39015910360548234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,51200,1.4885191387600367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,16384,0.1318311161465115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,12288,0.2476968765258789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,16384,0.35868178473578555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,51200,1.1937413745456271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,51200,0.9558853573269315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,12288,0.09879644711812337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,12288,0.2578053209516737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,128,0.014905777242448596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,10240,0.08439022302627563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,8192,0.1914151112238566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,10240,0.25672001308865017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,10240,0.2161760065290663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,8192,0.07362844546635945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,8192,0.17642577489217123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,7168,0.16369867324829102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,7168,0.06121777825885349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,10240,0.23216443591647676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,7168,0.1582355499267578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,6144,0.13338222768571642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,65536,2.0703858269585504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,5120,0.10091288884480794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,6144,0.054133332437939115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,6144,0.14126400152842203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,6144,0.13526933723025852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,7168,0.177656888961792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,5120,0.04630044433805677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,4096,0.08681599961386786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,4096,0.09574933184517755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,3584,0.07095288568072848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,4096,0.0389395554860433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,4096,0.09597600168652004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,3072,0.06977333625157674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,3584,0.035440001222822405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,3584,0.08458844158384536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,3584,0.0852924452887641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,3072,0.07432799869113497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,8192,0.1837599939770169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,3072,0.07737333244747586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,5120,0.11888089444902207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,2560,0.028708444701300725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,2560,0.06415733363893297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,2560,0.06545866860283746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,2048,0.05436266793145073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,2048,0.024175999893082514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,2048,0.05547022157245212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,2560,0.06176977687411838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,1536,0.04308266772164238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,1536,0.04552622305022346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,5120,0.11612533198462592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,2048,0.044624000787734985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,12288,0.2757279872894287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,1024,0.03246488836076524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,1024,0.03579911258485582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,1536,0.03499822152985467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,768,0.02665688925319248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,3072,0.031239999665154353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,512,0.01587199999226464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,768,0.01640888883007897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,768,0.0303439994653066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,512,0.015164444843928019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,512,0.02197066611713833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,256,0.011728888584507836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,512,0.02493688960870107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,1024,0.017832888497246634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,128,0.01036622209681405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,256,0.013777777552604675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,256,0.018929777873886954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,256,0.01993333299954732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,64,0.009376000199053023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,128,0.01755555636352963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,1024,0.02508977717823452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,64,0.018072000808186002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,32,0.01791111131509145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,128,0.014175999495718213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,1536,0.020018666982650757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,65536,1.184077368842231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,65536,0.44568533367580837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,51200,0.7816098001268176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,65536,1.577628453572591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,128,0.018222222725550335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,16384,0.28692711724175346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,65536,1.2188453674316406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,51200,0.35741954379611546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,16384,0.31292798784044057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,768,0.02075199948416816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,16384,0.1131955517662896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,16384,0.29768088128831655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,12288,0.08117600282033284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,10240,0.18336978223588732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,12288,0.2077102263768514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,32,0.009353777600659264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,10240,0.19325955708821616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,8192,0.14752266142103407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,10240,0.1829528941048516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,51200,1.295797348022461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,8192,0.056608888838026256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,7168,0.12939555115169948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,8192,0.14294311735365126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,12288,0.22955287827385795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,7168,0.0510480006535848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,7168,0.13499911626180014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,7168,0.12855554951561823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,10240,0.06953333483801948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,5120,0.09014488591088189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,6144,0.11726400587293838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,6144,0.10954755544662476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,12288,0.21459644370608857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,8192,0.1550640000237359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,5120,0.037231110864215426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,5120,0.09916088978449504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,5120,0.09494933154847886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,3584,0.061531556977166064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,4096,0.0802782244152493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,6144,0.11650221877627903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,4096,0.07835200097825792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,3072,0.05229422118928698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,3584,0.028730667299694482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,3584,0.07106488943099976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,6144,0.04378755556212532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,3584,0.07071022192637126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,2560,0.0414319998688168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,3072,0.025853332546022203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,3072,0.06249777475992838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,4096,0.07012622223960029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,3072,0.06279999679989286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,2560,0.02347022294998169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,2560,0.053893334335751004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,2560,0.05594933032989502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,2048,0.04509244362513224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,2048,0.046359111865361534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,1536,0.036899556716283165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,1024,0.01980622278319465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,51200,0.8987048467000326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,1024,0.02758311066362593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,1536,0.027590221828884546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,768,0.016226665841208566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,1024,0.029528889391157363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,1536,0.01807733376820882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,1536,0.03822222352027893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,512,0.012478222449620565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,768,0.025109334124459162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,2048,0.03421422176890903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,512,0.018925334016482036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,512,0.01314933349688848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,1024,0.01531377765867445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,512,0.021134222547213238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,256,0.012425777812798819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,128,0.008648888932334052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,256,0.018062222335073683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,768,0.014489778214030795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,128,0.012423999607563019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,64,0.007727999654081132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,128,0.01590755581855774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,256,0.009810667071077559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,4096,0.03183022141456604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,32,0.007925333248244392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,64,0.01550222271018558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,256,0.016294222739007738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,32,0.015556444724400839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,768,0.02273066673013899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,128,0.015605333778593274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,65536,0.9307439592149523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,65536,0.41357866923014325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,2048,0.020426667398876615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,51200,0.7023066414727105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,16384,0.2700177828470866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,51200,0.3274426725175646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,65536,1.3819733725653753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,65536,1.054599126180013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,16384,0.2861848937140571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,12288,0.20135111278957793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,51200,1.0673369301689994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,16384,0.2513066662682427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,12288,0.07793244388368395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,10240,0.17254222763909233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,12288,0.18034666114383271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,10240,0.07251377900441487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,8192,0.1272666719224718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,10240,0.17228621906704375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,10240,0.1539964410993788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,7168,0.12391022841135661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,8192,0.13760177294413248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,16384,0.11312977472941081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,7168,0.12327199512057835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,6144,0.10354577832751805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,7168,0.11158488856421576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,12288,0.22634310192532012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,5120,0.0912062194612291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,6144,0.10626666413413154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,6144,0.09886578056547378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,5120,0.034972445832358465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,5120,0.0888906651073032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,4096,0.06676355335447523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,8192,0.12683378325568304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,51200,0.8089395629035102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,7168,0.04773955543835958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,3584,0.05916533205244276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,4096,0.07056088580025567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,6144,0.04090844591458639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,3584,0.06503555509779188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,3584,0.026521777113278706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,3584,0.0630942251947191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,8192,0.052582220898734204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,3072,0.02369155486424764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,2560,0.04128444525930617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,3072,0.05636799997753567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,2560,0.02260977857642704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,2048,0.03487022386656867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,3072,0.0491555564933353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,2048,0.041033777925703264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,2048,0.018644443816608854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,5120,0.08362578021155463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,1536,0.023595554961098567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,4096,0.07212266657087538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,1536,0.03318488929006789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,2560,0.04890311095449659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,1024,0.016680000556839835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,1536,0.01643822259373135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,1024,0.014276444911956787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,1024,0.024676443801985845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,1024,0.026604443788528442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,3072,0.05727110968695747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,2048,0.04144355654716492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,768,0.020621332857343886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,768,0.02274311085542043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,4096,0.029838220940695867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,1536,0.034648001194000244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,512,0.012440000143316058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,512,0.019280888968043856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,768,0.01428711083200243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,256,0.014937778313954672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,128,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,256,0.012127111355463663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,256,0.016873778568373788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,768,0.013227555486891003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,2560,0.0488684442308214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,64,0.007287110719415877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,512,0.017492444978819955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,64,0.014273777604103088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,32,0.014173333843549093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,512,0.0118568887313207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,65536,0.7513173421223959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,128,0.011759999725553723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,65536,0.36531821886698407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,65536,1.084268463982476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,51200,0.664189338684082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,65536,1.0207547081841362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,32,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,16384,0.22844534450107148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,51200,0.9263653225368924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,51200,0.7649706734551324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,16384,0.2419813209109836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,16384,0.2496088875664605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,12288,0.1810328827963935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,128,0.015212444795502557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,12288,0.06924977567460802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,12288,0.1800746652815077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,16384,0.09091377920574611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,12288,0.16583732763926187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,10240,0.15272443824344212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,10240,0.1489644447962443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,8192,0.10807466506958008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,8192,0.12441155645582412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,10240,0.15330221917894152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,8192,0.12484089533487956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,7168,0.1015697783894009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,10240,0.05974133147133721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,128,0.014456889695591398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,7168,0.04161599940723843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,6144,0.09543733464346991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,6144,0.09570311175452338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,6144,0.03660710983806186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,5120,0.06625511248906453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,7168,0.10829066567950779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,5120,0.08139289087719388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,5120,0.03223022156291538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,5120,0.08193333281411065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,4096,0.06474400228924222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,4096,0.06799288590749104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,3584,0.04500088757938809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,4096,0.027467555469936792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,256,0.009303111169073317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,3584,0.0621395574675666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,3072,0.03832177652253045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,8192,0.04679110977384779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,3072,0.05011377731959025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,3072,0.022513777017593384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,2560,0.0326186650329166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,3072,0.053618666198518544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,51200,0.2994728883107503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,4096,0.05537066525883145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,2048,0.028402666250864666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,2560,0.020278222031063504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,2560,0.04679110977384779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,2048,0.03645244571897719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,2048,0.04032444291644626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,1536,0.02971733278698391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,3584,0.02516977820131514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,2560,0.04361244373851352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,1024,0.016173332929611206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,1536,0.0158942225906584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,2048,0.01773066653145684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,1536,0.02237244447072347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,768,0.012618666721714867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,1024,0.02619022296534644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,7168,0.10897066858079697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,768,0.018586667047606576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,512,0.01036444471942054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,768,0.022301332818137273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,1536,0.03278933299912347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,512,0.015613334046469795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,256,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,512,0.01891999940077464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,1024,0.02204622162712945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,1024,0.013743110828929476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,128,0.007283555964628856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,256,0.011424889167149862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,256,0.013620444469981723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,3584,0.05677511294682821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,64,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,128,0.011407111254003314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,32,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,128,0.013801777528391944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,64,0.012711111042234631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,32,0.013125333521101209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,512,0.01219733307758967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,65536,0.6603084670172797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,6144,0.08659555514653523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,65536,0.3098764419555664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,128,0.012787555654843649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,768,0.013066666821638743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,51200,0.5749351183573405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,16384,0.18175378110673693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,65536,0.8022551006740994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,16384,0.08116088973151313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,51200,0.761564466688368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,51200,0.6273360252380371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,16384,0.20406577322218153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,256,0.01519377695189582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,12288,0.06038666433758206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,12288,0.16021600034501818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,65536,0.9302728441026477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,51200,0.24973956743876138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,10240,0.1342328919304742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,8192,0.09737600220574273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,10240,0.12906755341423884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,16384,0.2407217820485433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,8192,0.04064355625046624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,7168,0.07934400108125475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,8192,0.10572977860768636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,10240,0.11694399515787761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,6144,0.06994933552212186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,7168,0.09600444634755452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,12288,0.1567013396157159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,7168,0.0965048869450887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,10240,0.05242044395870633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,6144,0.08361866739061143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,5120,0.05771733654869927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,6144,0.08243822389178805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,8192,0.10846222109264797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,5120,0.07060177458657159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,4096,0.04711822337574429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,5120,0.02817866537306044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,5120,0.07077511151631673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,3584,0.04049244523048401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,4096,0.024001777172088623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,4096,0.05702310800552368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,4096,0.058984888924492725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,6144,0.0331928895579444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,3584,0.05289599961704678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,12288,0.1364533371395535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,3072,0.04565511147181193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,2560,0.02953333324856228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,3072,0.020078221956888836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,3072,0.047464887301127114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,2048,0.02439555525779724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,2560,0.018398221996095445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,2560,0.041497776905695595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,2048,0.03330399923854404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,1536,0.01998044384850396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,2048,0.035403556293911405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,1536,0.027012444204754297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,1536,0.028753777345021565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,1024,0.014984889162911309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,7168,0.03632355398601956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,1024,0.019888000355826486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,2048,0.016153777639071148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,768,0.012151999606026543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,1024,0.012776888906955719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,768,0.016751110553741455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,768,0.01994222237004174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,1536,0.014475555883513557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,2560,0.03933066791958279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,512,0.01425333321094513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,1024,0.022868444522221882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,256,0.007669332954618666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,256,0.012148444851239523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,512,0.009905777871608734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,256,0.010651555326249866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,128,0.00665155549844106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,256,0.014909333652920194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,3584,0.022111111217074927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,512,0.016538666354285348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,64,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,128,0.011716444459226398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,32,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,64,0.011399110986126794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,768,0.012088889049159156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,32,0.011526222030321756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,3072,0.033956444925732084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,65536,0.541870223151313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,128,0.01074044406414032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,128,0.01348444405529234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,65536,0.2811173333062066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,51200,0.4556728998819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,65536,0.6659679942660862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,65536,0.7752488984002007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,16384,0.14808532926771376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,51200,0.2263368765513102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,51200,0.5931155416700575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,16384,0.19257956080966523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,16384,0.16734577549828422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,51200,0.5146649148729112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,12288,0.13703021738264295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,12288,0.12576533688439265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,10240,0.11555466387006973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,16384,0.08709244595633613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,3584,0.05089866783883837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,512,0.01129155523247189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,10240,0.09755733278062607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,8192,0.09312711159388225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,7168,0.06528088781568739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,8192,0.08788533343209161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,7168,0.08208622535069783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,8192,0.07752622498406304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,7168,0.07840089003245036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,6144,0.05830666753980848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,12288,0.05788355403476291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,6144,0.07111199696858723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,8192,0.03804355528619554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,5120,0.044453332821528115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,5120,0.06003733476003011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,10240,0.10655911101235284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,5120,0.0592213339275784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,4096,0.03789688812361823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,5120,0.026439110438028973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,7168,0.03393777873780992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,3584,0.03203644355138143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,4096,0.022229333718617756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,4096,0.04912622107399834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,4096,0.05017155408859253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,6144,0.07058933046129015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,3584,0.04418311185306973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,3584,0.04492533206939697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,12288,0.11542844772338867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,3072,0.017808889349301655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,2560,0.02330133318901062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,6144,0.03040800160831875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,3072,0.0395706660217709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,2560,0.03466399841838413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,2048,0.019401777121755812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,2560,0.01670222149954902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,2048,0.02843911117977566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,2048,0.02989333205752903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,2560,0.03404711021317376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,2048,0.015670221712854173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,10240,0.05022577775849236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,1536,0.013099555340078143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,1536,0.024791111548741657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,3072,0.039975110027525164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,1024,0.017606221967273288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,768,0.009697777529557547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,3072,0.027296000056796606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,1536,0.015392889579137167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,768,0.014871999621391296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,512,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,1536,0.02351288828584883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,768,0.01719466679626041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,1024,0.011750222080283694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,256,0.006565333240562015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,512,0.010471999645233154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,512,0.014512888259357877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,1024,0.019612444771660697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,256,0.011000889043013254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,768,0.011019555230935415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,128,0.01035644445154402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,128,0.009874666730562845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,128,0.012436444560686747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,1024,0.01143200033240848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,32,0.005614222337802251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,64,0.010095111197895473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,32,0.01034400032626258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,65536,0.42831465933058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,512,0.012538666526476542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,64,0.005296000176005893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,65536,0.6100302272372776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,51200,0.2851457860734728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,256,0.013223999904261695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,65536,0.6474764611985948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,51200,0.18781067265404594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,16384,0.11302044656541611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,51200,0.479289796617296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,51200,0.4922151035732693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,16384,0.15209600660536024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,12288,0.08072977595859103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,16384,0.16405599647098118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,65536,0.23334754837883842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,128,0.005583111196756363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,10240,0.07716888851589628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,12288,0.12478489345974392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,10240,0.04279288980695936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,16384,0.061736000908745654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,8192,0.06335466437869601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,256,0.009744000103738572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,8192,0.07795555724038018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,7168,0.055607113573286265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,12288,0.049590223365359835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,10240,0.09716266393661499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,10240,0.10540711217456394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,7168,0.02921688887808058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,3584,0.020459555917316012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,6144,0.05968177980846829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,12288,0.11481244034237331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,5120,0.038429333104027644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,6144,0.06725777520073785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,7168,0.06850044594870673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,7168,0.07825066645940144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,5120,0.04981777734226651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,5120,0.02260444396071964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,5120,0.057915555106268994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,4096,0.019655999210145738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,8192,0.08691200282838608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,8192,0.0321742229991489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,6144,0.02602133320437537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,3584,0.03756977783309089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,3072,0.021038222644064162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,3584,0.04386311107211643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,6144,0.044454223579830594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,3072,0.01681866745154063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,2560,0.018573333819707234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,4096,0.04798310995101929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,2560,0.028455999162462022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,3584,0.01794755624400245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,2048,0.01589777734544542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,4096,0.027277333868874445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,3072,0.03306844499376085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,4096,0.04137066668934292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,2048,0.014211555322011312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,1536,0.020240889655219186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,2560,0.015421332584487068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,1536,0.012433778080675336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,1024,0.009735110733244155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,2048,0.024153777294688757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,3072,0.0388648874229855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,3584,0.024704888463020325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,1024,0.015202666322390238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,768,0.008368000388145447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,1024,0.019302222463819716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,2560,0.03340977761480544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,768,0.013148444394270578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,768,0.010560000108347999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,768,0.01682044400109185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,512,0.011069333387745751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,1536,0.012862222062216865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,256,0.006151999864313338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,512,0.010064000056849586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,1024,0.01108088923825158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,256,0.009697777529557547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,256,0.009680888719028896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,256,0.012028444144460889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,512,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,1536,0.02420088814364539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,128,0.008997333546479544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,64,0.00462755560874939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,32,0.0047208890318870544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,2048,0.029025776518715754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,64,0.00868622213602066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,32,0.009031111167536842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,65536,0.2948239909278022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,128,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,51200,0.22857244809468588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,65536,0.18548799885643852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,65536,0.46986844804551864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,128,0.009457777771684859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,65536,0.48287020789252383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,128,0.011397333608733283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,51200,0.36677156554328066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,512,0.014190221826235453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,51200,0.3977546691894531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,12288,0.05910488631990221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,16384,0.12075377835167779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,16384,0.05128177669313219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,10240,0.04807022213935852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,12288,0.04049422343571981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,12288,0.09066133366690741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,8192,0.038789331912994385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,10240,0.03519466519355774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,10240,0.07647022273805407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,51200,0.1482951111263699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,8192,0.02626488937271966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,8192,0.06853510936101277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,16384,0.12730311022864446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,7168,0.023592889308929443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,12288,0.09898489051394993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,16384,0.08281689220004611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,8192,0.06143289142184787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,6144,0.021577777134047613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,5120,0.024619556135601465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,7168,0.054429332415262856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,7168,0.06115111377504137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,6144,0.02716444432735443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,5120,0.040511998865339495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,4096,0.020020445187886555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,5120,0.01870044403605991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,10240,0.08328621917300753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,4096,0.016365332735909354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,7168,0.03282666537496779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,4096,0.0392177767223782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,6144,0.0536702209048801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,3072,0.01569955547650655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,3072,0.026515555050637987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,3584,0.01776266594727834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,3072,0.031143112315071955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,3072,0.014358222484588623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,3584,0.030304888884226482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,3584,0.034761776526769005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,2560,0.023128888673252527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,2048,0.011792889071835412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,2560,0.0274008893304401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,6144,0.04759111007054647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,2048,0.023535999986860488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,2560,0.013766222529941134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,1536,0.009736888938479954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,3584,0.015354666444990369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,1536,0.011071110765139261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,2560,0.01312711089849472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,1024,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,1536,0.019766221443812054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,5120,0.04617866542604235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,768,0.0069884442620807225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,1024,0.010038221875826517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,1024,0.012789333032237159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,1024,0.016200888488027785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,512,0.00591822216908137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,2048,0.011971555650234222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,768,0.014478221535682678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,512,0.009424000150627559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,1536,0.016576889488432143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,256,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,2048,0.01995022263791826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,768,0.011077333655622272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,256,0.00870311094654931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,128,0.004237333519591225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,256,0.011022222538789114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,512,0.009349333743254343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,128,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,512,0.012427555190192329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,128,0.00867733359336853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,32,0.0038888889054457345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,64,0.007346666521496243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,32,0.007640889121426477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,768,0.009748444788985783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,65536,0.2419813209109836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,256,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,64,0.003939555750952827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,65536,0.4142684406704373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,128,0.010656000011497073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,51200,0.1912400060229831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,16384,0.06545244322882758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,65536,0.4823919932047526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,4096,0.03353066576851739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,51200,0.3261360062493218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,16384,0.050536000066333346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,12288,0.04481866624620226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,16384,0.11002400186326768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,16384,0.1272373331917657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,10240,0.03990755478541056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,12288,0.039333333571751915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,65536,0.18398222658369276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,12288,0.08174222045474581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,8192,0.030812445614073012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,12288,0.09739555252922906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,10240,0.03439911206563314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,7168,0.027319111757808264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,10240,0.06857866711086698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,8192,0.025980444418059453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,10240,0.08223644230100843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,51200,0.1472053395377265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,6144,0.023810666468408372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,8192,0.06827555762396918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,7168,0.049161778555976025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,5120,0.02036622166633606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,6144,0.04269244604640537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,5120,0.03649155630005731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,5120,0.018616000811258953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,5120,0.04590310984187656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,4096,0.016951110627916124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,7168,0.02330577704641554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,7168,0.06048622396257189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,3584,0.015249777171346875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,6144,0.05249155561129252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,3584,0.0273964438173506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,3584,0.03373688790533278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,8192,0.05535466803444756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,4096,0.030221333106358845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,3072,0.024421332610978022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,3072,0.014440889159838358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,2560,0.011980444192886353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,3072,0.03045066528850132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,3584,0.015379556351237826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,2560,0.021271111236678228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,3072,0.013821333646774292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,2048,0.010398222340477837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,2560,0.026643556025293138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,2560,0.013150222599506378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,1536,0.00869777798652649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,6144,0.02126400007141961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,2048,0.01829155617290073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,2048,0.023039999935362074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,1024,0.007322666545708974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,1536,0.019301333361201815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,1024,0.011800888511869641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,1024,0.010093332992659675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,768,0.006225777582989798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,51200,0.3786142137315538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,4096,0.03785955574777391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,768,0.010399999717871347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,1536,0.011086222198274402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,768,0.014481777946154276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,512,0.009015111459626092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,1024,0.015617777903874716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,256,0.004581333448489507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,4096,0.01647644407219357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,1536,0.015255999233987598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,256,0.011071999867757162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,256,0.009032888544930352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,512,0.012420444852775998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,2048,0.01204711116022534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,128,0.008714666797055138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,256,0.007449777589903937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,64,0.0038773334688610504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,32,0.0038773334688610504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,128,0.003929777691761653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,64,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,65536,0.1778728829489814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,32,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,512,0.009205333060688442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,51200,0.14224355750613743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,65536,0.13879644870758057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,65536,0.33684621916876895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,128,0.010678222609890832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,512,0.0052435555391841466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,16384,0.04660177893108792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,51200,0.11151022381252712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,51200,0.2651546796162923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,51200,0.2602906756930881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,16384,0.08860444360309178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,12288,0.0344782206747267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,128,0.007027555671003129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,16384,0.09011644124984741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,10240,0.026075555218590632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,12288,0.03137600090768602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,12288,0.06730222039752536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,12288,0.0692888895670573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,768,0.009706666900051964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,8192,0.02125777800877889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,10240,0.056519110997517906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,10240,0.059559113449520536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,7168,0.017966222431924608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,8192,0.04632444514168633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,16384,0.039290666580200195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,7168,0.01792533364560869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,6144,0.015556444724400839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,7168,0.04384266667895847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,10240,0.02774311105410258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,6144,0.03882222374280294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,6144,0.016391999191708036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,5120,0.01367644468943278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,8192,0.04928177926275465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,4096,0.01127199994193183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,5120,0.030604445272021826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,7168,0.04067555401060317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,5120,0.03382577829890781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,3584,0.010386666489972008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,4096,0.01313688854376475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,4096,0.025410667061805725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,4096,0.02831377916865879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,65536,0.33167823155721027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,3072,0.009391999906963771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,3584,0.025690666503376428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,3072,0.022988445229000513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,3072,0.011573333707120685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,2560,0.008345777789751688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,2560,0.017954667409261067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,6144,0.03592977921168009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,3584,0.02275644408331977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,2560,0.011074666347768573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,8192,0.01966399947802226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,2048,0.015465777781274585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,2048,0.010399111443095738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,2048,0.018001778258217704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,5120,0.014527110589875115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,1536,0.009740444521109263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,1536,0.012859555582205454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,1536,0.015596444408098856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,2560,0.020861332615216572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,3584,0.012092444631788464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,768,0.004908444566859139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,1024,0.010383110907342698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,1024,0.012693333129088083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,768,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,512,0.004561777744028303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,768,0.011767110890812345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,1024,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,512,0.010402667025725046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,2048,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,1024,0.0090506664580769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,256,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,3072,0.02034222251839108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,768,0.008731555607583787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,256,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,512,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,128,0.009015999734401703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,512,0.008336000144481659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,64,0.006263111200597551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,1536,0.006587555425034628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,128,0.0035991109907627106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,256,0.009727111293209923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,65536,0.10506311390135024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,32,0.006314666734801398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,128,0.006355555521117316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,51200,0.08783466948403253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,64,0.003314666656984223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,65536,0.29317066404554576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,65536,0.3169368902842204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,51200,0.11067110962337917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,32,0.003543111185232798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,51200,0.2308808962504069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,51200,0.24889779090881348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,12288,0.024485333098305598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,16384,0.03894933395915561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,16384,0.07680710819032457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,16384,0.08556711011462742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,10240,0.020498666498396132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,12288,0.030585779084099665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,65536,0.1375919977823893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,12288,0.06616089079115126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,8192,0.01740800009833442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,10240,0.026533333791626826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,10240,0.04979022343953451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,16384,0.031714667876561485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,10240,0.056688891516791455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,7168,0.015226667126019796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,8192,0.019319110446506076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,8192,0.04025333457522922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,8192,0.047028445535235934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,6144,0.013555555707878537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,7168,0.041839109526740186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,256,0.0038897775941424896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,6144,0.031488888793521456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,6144,0.016195555528004963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,6144,0.036545776658587985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,5120,0.026774222652117412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,5120,0.014504000544548035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,4096,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,128,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,5120,0.012071111136012606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,4096,0.012772444221708508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,3584,0.008346666892369589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,7168,0.035952889257007174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,3584,0.02029244436158074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,12288,0.05823377768198649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,3072,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,3584,0.02392888896995121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,4096,0.022364444202846948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,2560,0.00674933319290479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,3072,0.011620444556077322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,4096,0.026603556341595117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,3072,0.021661332911915247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,2048,0.006046222315894232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,2560,0.01110311100880305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,2560,0.01958755486541324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,5120,0.031699554787741765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,1536,0.0053004444473319584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,2048,0.014021333720948962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,2048,0.01036888857682546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,2048,0.017231111725171406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,1024,0.004569777597983678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,1536,0.011740444435013665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,1536,0.014519999424616495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,3072,0.018262222409248352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,768,0.004179555508825514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,1024,0.0094355551732911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,1024,0.00905688852071762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,1024,0.012325333224402534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,512,0.0038862224254343244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,768,0.008402666283978356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,7168,0.017635555730925668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,768,0.011415111521879831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,3584,0.012509333590666452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,512,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,512,0.007303111255168915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,1536,0.009704888694816166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,128,0.003269333392381668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,256,0.007978666987684038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,256,0.006046222315894232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,256,0.00906577789121204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,64,0.0032071111102898917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,128,0.007689778175618913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,32,0.0032257777121331957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,768,0.008678221868144141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,64,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,256,0.0033653332955307434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,512,0.009782222410043081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,65536,0.252345773908827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,65536,0.13726400004492864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,128,0.005624889085690181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,2560,0.016134222348531086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,65536,0.07650044229295519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,32,0.005635555419656966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,51200,0.24630400869581434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,16384,0.06643733051088122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,51200,0.06334044535954793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,16384,0.08453333377838135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,65536,0.3134044541252984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,16384,0.040172444449530706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,12288,0.03176711002985636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,51200,0.1979964441723294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,10240,0.015547555353906421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,12288,0.06531466378106011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,10240,0.02769600020514594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,8192,0.013350221845838757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,10240,0.05581777625613742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,12288,0.017669333351982962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,128,0.008459555606047312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,7168,0.012283555335468717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,8192,0.02035466664367252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,12288,0.050683554675843984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,16384,0.027474666635195415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,7168,0.03107200066248576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,7168,0.041152887874179415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,10240,0.04265333215395609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,6144,0.01626399987273746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,5120,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,6144,0.03587999939918518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,8192,0.034974220726225115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,8192,0.04607111215591431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,5120,0.015171556009186639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,51200,0.11008889145321316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,5120,0.031324442889955305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,6144,0.010545777777830759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,3584,0.007299555672539606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,4096,0.01983022193113963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,4096,0.026439110438028973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,6144,0.027013333307372198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,3584,0.01791555517249637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,3072,0.006959111326270633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,3584,0.023800000548362732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,5120,0.023588443795839947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,3072,0.016182222300105624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,2560,0.005971555494599872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,7168,0.018400889303949144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,4096,0.013470222552617392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,2560,0.014465777410401238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,2048,0.005532444351249271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,3584,0.012786666552225748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,4096,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,2048,0.01697688963678148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,2048,0.010544000400437249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,1536,0.004899555610285865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,2560,0.011181332998805575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,2560,0.019275556008021038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,1536,0.014487110906177096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,1536,0.009893333746327294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,1024,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,1024,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,768,0.0039048890272776284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,1024,0.012097777591811286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,3072,0.011777777638700275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,3072,0.021528888079855178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,768,0.008683555656009251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,768,0.00798044436507755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,768,0.011390222443474663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,256,0.0033688888781600525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,512,0.00833777752187517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,512,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,1024,0.004234666625658671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,128,0.003177777760558658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,256,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,256,0.005663110978073544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,256,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,2048,0.012515555653307172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,128,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,32,0.0031893334041039148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,512,0.003871110992299186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,64,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,65536,0.07197155555089314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,51200,0.057578669653998486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,1536,0.010753778119881948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,16384,0.024104000793562994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,512,0.00981866651111179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,65536,0.2596204545762804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,12288,0.014111111561457316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,10240,0.01256977766752243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,51200,0.19787200291951498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,128,0.007720888488822513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,8192,0.011061333119869232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,7168,0.0107351111041175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,12288,0.05161688725153605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,6144,0.010399111443095738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,10240,0.04314400090111626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,8192,0.034684442811542086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,5120,0.008990222381220924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,4096,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,6144,0.02677333354949951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,3584,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,4096,0.019633778267436557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,64,0.0028684443483750024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,16384,0.06631288925806682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,2560,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,3072,0.01623644431432088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,2048,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,2560,0.014507555299335055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,32,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,1536,0.004898666507667965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,1024,0.003912000192536248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,2048,0.012657778130637275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,768,0.003535110917356279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,1024,0.00872622181971868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,3072,0.006604444649484422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,3584,0.01789066692193349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,768,0.007703111403518253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,512,0.0069857777820693115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,128,0.002847111059559716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,256,0.005616888817813661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,64,0.002862222285734283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,5120,0.02347999976740943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,128,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,32,0.003120888852410846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,64,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,65536,0.07637421952353583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,512,0.003569777641031477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,51200,0.058562669489118785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,32,0.005245333330498801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,256,0.0032142222755485107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,12288,0.010773333410422007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,65536,0.2522266705830892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,7168,0.030682666434182063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,16384,0.06745866934458415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,8192,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,12288,0.05145600106981066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,7168,0.009703111317422655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,1536,0.01072266697883606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,8192,0.03439822130733066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,6144,0.009008000294367472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,16384,0.023360000716315374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,5120,0.011023999916182624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,6144,0.026760000321600173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,10240,0.010028444230556488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,5120,0.023542222049501207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,3584,0.008685333033402761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,10240,0.04234222239918179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,3072,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,7168,0.0306346681382921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,2560,0.007305777735180325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,3072,0.016252444850073922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,2048,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,51200,0.19797421826256645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,2560,0.014381332529915703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,1024,0.004259555704063839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,2048,0.012719111310111152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,1536,0.010773333410422007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,768,0.0038764443662431505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,4096,0.019713777634832595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,1024,0.0086666668454806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,256,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,768,0.007672888537247975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,3584,0.01814133259985182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,256,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,64,0.002879111096262932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,128,0.0052800000541739995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,32,0.0029093333416514923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,64,0.005116444494989184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,512,0.003363555504216088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,32,0.005285333428117964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,128,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,4096,0.009637333452701569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,512,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,1536,0.004562666846646203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,16384,1.2845502429538302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,16384,3.949706607394748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,12288,2.8386720021565757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,16384,4.70522944132487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,12288,1.0100079642401802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,51200,5.095439063178168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,10240,0.8438106642829047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,12288,3.222863939073351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,10240,2.760504828559028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,8192,0.6496906810336643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,8192,2.486917283799913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,10240,2.6063512166341147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,7168,0.5826746622721354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,7168,1.7630799611409504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,8192,2.0422560373942056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,6144,0.502778689066569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,6144,1.7785199483235676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,6144,1.5012640423244898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,51200,12.354659186469185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,5120,0.43946221139695907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,7168,1.8666790856255426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,5120,1.3241555955674913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,4096,0.3585973315768772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,4096,1.0997848510742188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,3584,0.8510453436109754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,3584,0.319160885281033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,3584,0.8465369012620715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,4096,0.9588417477077908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,3072,0.28487732675340444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,3072,0.8313733206854926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,5120,1.241893344455295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,3072,0.734332455529107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,2560,0.2521173424190945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,2560,0.6090755462646484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,2048,0.21188178327348497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,2048,0.5303520096672906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,2048,0.4986257553100586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,1536,0.4477715492248535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,1536,0.16861511601342094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,1024,0.2961964342329237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,1536,0.3948497772216797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,1024,0.15237332714928523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,2560,0.6287733183966743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,512,0.16215556197696263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,768,0.2548942301008436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,256,0.1156560050116645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,512,0.20137600104014078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,256,0.0916702217525906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,768,0.21154133478800455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,128,0.08711910910076565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,1024,0.2953520086076525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,768,0.1143946647644043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,64,0.07398666938145955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,128,0.08947377734714085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,32,0.07832622528076172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,128,0.1706026660071479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,512,0.11122222741444905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,64,0.17483199967278373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,32,0.17666844526926676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,256,0.18111111058129203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,51200,18.31624518500434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,128,0.00833777752187517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,16384,3.1232310401068797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,16384,3.4241786532931857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,51200,3.9841893513997397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,16384,1.0066133075290256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,65536,5.099701351589627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,12288,0.7532551023695203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,12288,2.6485769483778214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,12288,2.5895777808295355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,10240,0.709952884250217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,51200,9.702704535590277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,10240,2.1761101616753473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,8192,1.5299243927001953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,10240,2.008509318033854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,8192,0.5083786646525065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,7168,0.44593509038289386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,8192,1.5823137495252821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,65536,12.171947055392794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,6144,0.39558667606777614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,6144,1.3555439843071833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,6144,1.1498799853854709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,7168,1.2902000215318468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,5120,0.3338453239864773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,7168,1.3407546149359808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,4096,0.7195662392510308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,5120,1.0730240080091689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,4096,0.2760515477922228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,4096,0.7449350886874728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,3584,0.6697697639465332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,3584,0.24761599964565703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,3072,0.6313680013020834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,51200,14.231874254014757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,3584,0.671471118927002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,3072,0.567305776807997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,5120,0.9497884114583334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,2560,0.19769956005944145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,2048,0.41573066181606716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,2048,0.16295822461446127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,1536,0.33606844478183323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,2048,0.39232977231343585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,1536,0.13110844294230142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,1024,0.2452275488111708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,1536,0.32865956094529897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,3072,0.21792533662584093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,768,0.1877866718504164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,2560,0.5365262031555176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,1024,0.23268445332845053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,512,0.13457243972354466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,768,0.18540888362460664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,512,0.08358310990863377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,2560,0.5446773105197483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,512,0.15886221991644964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,128,0.06752000252405803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,256,0.14306399557325575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,128,0.07203021976682875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,64,0.05679021941290962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,128,0.13464532958136663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,32,0.05947022305594551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,768,0.0913706686761644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,64,0.13791999551984999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,256,0.09535200066036648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,256,0.07378577523761325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,1024,0.102801779905955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,32,0.1398764451344808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,65536,19.274478488498264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,65536,1.6097911198933919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,16384,1.0235377417670357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,51200,1.205550193786621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,16384,0.41074132919311523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,16384,1.1183226903279622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,51200,3.0105306837293835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,12288,0.7547751002841525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,10240,0.6251982053120931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,12288,0.8265484703911675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,10240,0.6546862390306261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,8192,0.5240249103969997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,8192,0.177328003777398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,8192,0.5057697825961643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,7168,0.4703342119852702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,65536,3.9474143981933594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,10240,0.2110204431745741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,7168,0.1543413268195258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,6144,0.3619653383890788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,51200,4.832017686631945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,6144,0.1343475580215454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,5120,0.35027111901177305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,5120,0.1271866692437066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,4096,0.2738533284929064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,12288,0.3177226649390327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,4096,0.10691289107004802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,3584,0.22912089029947916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,4096,0.26774488555060494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,3072,0.19050221972995332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,3584,0.2217031055026584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,7168,0.4405831231011285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,3072,0.07932266924116346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,3072,0.19150755140516493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,2048,0.13280533419715032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,2560,0.16205867131551108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,65536,6.192833794487847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,2048,0.058465778827667236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,2048,0.13444711102379694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,3584,0.0881955557399326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,1536,0.04753333330154419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,1024,0.0768995549943712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,6144,0.3743564552730984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,1024,0.08151555723614164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,1024,0.039481778939565025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,768,0.05472444494565328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,5120,0.32146043247646755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,1536,0.0920320020781623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,512,0.04076355695724487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,768,0.06577333476808336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,512,0.03363644414477878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,1536,0.10665688912073772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,2560,0.06691022051705255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,256,0.05028444528579712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,128,0.02525955604182349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,256,0.030438221163219873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,64,0.021732444564501446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,128,0.029547555579079524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,128,0.047394666406843394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,768,0.03658488723966811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,64,0.048271109660466514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,256,0.031558222240871854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,2560,0.1604479948679606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,32,0.04890311095449659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,32,0.02515999972820282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,512,0.056009776062435575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,65536,1.1117804845174153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,51200,0.9159431457519531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,16384,0.7998879750569662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,16384,0.285500446955363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,12288,0.6270560158623589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,12288,0.19087910652160645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,16384,0.8222755326165093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,51200,2.5150843726264105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,65536,3.067360983954536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,12288,0.749443531036377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,10240,0.16456800036960179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,10240,0.6091938018798828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,8192,0.4383022255367703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,8192,0.1317324505911933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,10240,0.4857021967569987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,7168,0.38285332255893284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,7168,0.12353332837422688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,8192,0.4032035668690999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,6144,0.10526044501198663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,51200,3.509973314073351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,5120,0.2474151187472873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,6144,0.28585510783725315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,5120,0.08887377712461685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,4096,0.20643556118011475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,5120,0.23938311470879448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,4096,0.0735608869128757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,3584,0.17125156190660265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,3072,0.14349599679311117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,3584,0.17326933807796904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,3072,0.05870755513509115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,2560,0.1229413350423177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,3072,0.14740799532996282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,7168,0.3545706537034776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,2048,0.0972577797041999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,2560,0.1253688865237766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,2048,0.04596266812748379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,1536,0.07853955692715116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,2048,0.10403377479977077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,65536,4.378964318169488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,1024,0.05970222420162625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,3584,0.06646755668852064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,768,0.044488890303505786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,1024,0.0318524440129598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,6144,0.3218782212999132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,768,0.029453333881166246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,768,0.050438221957948476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,2560,0.053616000546349414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,256,0.024373332659403484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,512,0.02641155487961239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,512,0.04336355461014641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,128,0.020204444726308186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,256,0.02390844457679325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,4096,0.19303288724687365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,256,0.038763556215498186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,1536,0.03849155704180399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,64,0.01771022213829888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,128,0.023763555619451735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,32,0.017902221944597032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,128,0.03671111000908746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,1024,0.0631928907500373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,64,0.03703644540574815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,512,0.03471733464135064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,32,0.03761955433421665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,16384,0.6361262003580729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,1536,0.08293600214852227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,51200,0.7303911315070258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,16384,0.21681866380903456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,12288,0.5893857744004991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,51200,1.9198985629611547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,12288,0.49463558197021484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,12288,0.19493867291344535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,65536,0.9542755550808377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,10240,0.4564017719692654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,10240,0.14275377326541477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,8192,0.3872835636138916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,10240,0.40856088532341855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,16384,0.7164062394036187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,8192,0.11569421821170384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,7168,0.33189956347147626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,65536,3.9184739854600696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,65536,2.680845260620117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,6144,0.284313784705268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,7168,0.28574222988552517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,6144,0.08831555313534206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,5120,0.21432088481055367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,6144,0.2535884380340576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,51200,2.9951903025309243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,5120,0.20496266418033174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,8192,0.357735104031033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,4096,0.06515377759933472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,3584,0.14233867327372232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,4096,0.16211999787224665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,3072,0.11976710955301921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,3584,0.14347644646962485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,3072,0.05065777897834778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,2560,0.10223644309573704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,3072,0.1250444385740492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,2048,0.08527288834253947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,2560,0.04539555642339918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,5120,0.0756817791197035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,2560,0.10684977637396918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,1536,0.0656319989098443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,2048,0.03950400153795878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,1024,0.045927110645506114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,1536,0.03250577714708116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,768,0.036008887820773654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,1536,0.07037778033150567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,3584,0.05821333328882853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,1024,0.05336088935534159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,768,0.02424977719783783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,768,0.04274222254753113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,512,0.02204088866710663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,512,0.035756445593304105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,4096,0.19162310494316948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,128,0.01723022262255351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,256,0.020082667469978333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,256,0.031535999642478094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,2048,0.08955466747283936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,128,0.019869332512219746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,32,0.014894222219785055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,512,0.029810667037963867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,64,0.03020355436537001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,32,0.030104888810051814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,256,0.020672889219390023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,64,0.014682667122946845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,7168,0.10259822342130874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,65536,0.818393760257297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,65536,2.0278292761908636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,16384,0.6097920205858018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,51200,1.856312010023329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,1024,0.02665333284272088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,16384,0.1794888840781318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,12288,0.42859376801384824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,12288,0.13855377833048502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,16384,0.5660035345289443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,10240,0.3115075429280599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,12288,0.3957173294491238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,65536,3.1473253038194446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,51200,2.2572960323757596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,8192,0.25571733050876194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,8192,0.10318666696548462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,128,0.029934220843844946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,8192,0.2612248791588677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,7168,0.08539999855889215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,6144,0.21132889058854845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,7168,0.23795021904839408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,6144,0.07328889105055067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,6144,0.19810756047566733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,4096,0.12930133607652453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,5120,0.16789689328935412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,10240,0.33460532294379336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,4096,0.052942223019070096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,3584,0.1311724450853136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,3584,0.04804089003139072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,3584,0.11909422609541152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,5120,0.17926400237613252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,3072,0.041927112473381885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,10240,0.12045333120557998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,3072,0.10387733247545029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,7168,0.22454754511515299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,2560,0.03753333290417989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,2048,0.07183377610312568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,4096,0.13316267066531712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,1536,0.05347644289334615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,2048,0.032320890161726214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,2048,0.0733484427134196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,3072,0.09958666563034058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,1536,0.05865777863396538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,5120,0.06309333112504747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,768,0.02702399922741784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,1024,0.04443200098143684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,768,0.020656888683636982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,2560,0.08816533618503147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,768,0.03654222355948554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,512,0.029615110821194116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,512,0.018830221560266282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,256,0.01740444368786282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,2560,0.08735911051432292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,256,0.01718844473361969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,256,0.025956445270114478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,1536,0.02689244515366024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,128,0.025015999873479206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,64,0.011734222372372946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,128,0.017248888810475666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,32,0.012084444363911947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,1024,0.022675555613305833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,64,0.02474133339193132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,512,0.02119111186928219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,32,0.02456355591615041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,1024,0.03391822179158529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,51200,0.6259786817762587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,65536,0.6821857558356391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,51200,0.5316995514763726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,16384,0.4814000129699707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,51200,1.337270206875271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,65536,2.000044504801432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,12288,0.35547200838724774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,16384,0.5860373179117838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,16384,0.15698933601379395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,12288,0.1205822229385376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,10240,0.26730667220221627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,51200,2.135800043741862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,65536,2.664454142252604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,8192,0.25268978542751735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,10240,0.2918791241115994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,8192,0.08286755614810519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,7168,0.22253333197699654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,8192,0.23540356424119738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,6144,0.186280886332194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,7168,0.2040639983283149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,6144,0.06635822190178765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,6144,0.17602845033009848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,12288,0.3671591016981337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,4096,0.11858133474985759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,128,0.013845333622561561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,4096,0.047467556264665395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,3584,0.09681510925292969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,7168,0.07517155673768786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,3584,0.04174222217665779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,3072,0.09166844685872395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,3584,0.10564533207151626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,5120,0.14867999818589953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,2560,0.07103200091256036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,3072,0.03705155518319871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,5120,0.056891553931766085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,2560,0.03346488873163859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,2560,0.07963999774720934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,1536,0.04682488905058967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,4096,0.11971288257175022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,1536,0.024375999967257183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,1024,0.031055112679799397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,3072,0.09222755829493205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,1024,0.0400257772869534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,5120,0.14824977185991076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,1024,0.020452444752057392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,2048,0.028813332319259644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,2048,0.06574310859044392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,768,0.03260088960329691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,512,0.01702044407526652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,512,0.026714666022194758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,2048,0.0587973329755995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,10240,0.10196533468034531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,128,0.01276533305644989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,256,0.01586666703224182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,512,0.018382221460342407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,128,0.015858666764365304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,64,0.011199110911952125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,32,0.012464000119103326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,1536,0.05311555663744608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,256,0.014461333552996317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,64,0.02237777743074629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,32,0.02204177776972453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,256,0.02309422194957733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,768,0.01867111192809211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,128,0.02234755622016059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,16384,0.4101688861846924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,65536,1.7106089062160914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,51200,1.3208844926622179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,768,0.02344266739156511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,16384,0.17869067192077637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,12288,0.343615108066135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,16384,0.44671201705932617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,12288,0.11686666806538899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,10240,0.2427742216322157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,8192,0.19580177466074625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,51200,0.5377573437160915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,8192,0.07577600081761678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,8192,0.21287911468082002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,7168,0.18243911531236437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,65536,0.6863244374593099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,65536,2.332345750596788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,7168,0.06770310799280803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,6144,0.15550843874613443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,10240,0.09813155730565389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,6144,0.059383114178975425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,5120,0.1382933325237698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,6144,0.1542124483320448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,5120,0.05158133308092753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,5120,0.13050222396850586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,51200,1.671868430243598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,4096,0.1046746638086107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,3584,0.0977333320511712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,7168,0.17920800050099692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,3584,0.03785422113206651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,10240,0.25786489910549587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,3072,0.03462222218513489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,4096,0.10831911034054226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,3072,0.08097066481908162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,2560,0.07002400027381049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,2048,0.057651552889082164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,2560,0.0306346681382921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,2560,0.06959288650088839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,2048,0.025966222087542217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,2048,0.057709336280822754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,3072,0.08129600021574232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,1536,0.046775112549463906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,1024,0.01862044466866387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,1024,0.03523200088077121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,4096,0.04240177737341987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,12288,0.3089280128479004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,512,0.017408889200952318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,768,0.016908443636364408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,1536,0.022127999199761286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,512,0.01574133336544037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,3584,0.09250133567386204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,256,0.015505777464972602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,128,0.011065777805116443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,256,0.014750222365061441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,768,0.022082666556040447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,64,0.009746666583749983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,128,0.014767110347747803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,32,0.010066666536860997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,1536,0.04012444284227159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,64,0.01996888882584042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,32,0.01964977714750502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,1024,0.027190221680535212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,512,0.023806222611003455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,256,0.020664888951513503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,65536,0.5668373107910156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,768,0.02864266766442193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,65536,1.4547404183281794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,16384,0.33849512206183535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,16384,0.1495306624306573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,12288,0.2653048833211263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,16384,0.35508089595370823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,65536,1.9330533345540364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,12288,0.10744977659649319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,10240,0.2070524427625868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,128,0.0199608885579639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,12288,0.27242665820651585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,8192,0.16358221901787653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,10240,0.09415377510918511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,7168,0.14986133575439453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,51200,1.1425519519382052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,8192,0.1759528848859999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,6144,0.11570933130052354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,7168,0.060532444053226046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,7168,0.15375022093454996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,5120,0.11341867181989883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,6144,0.05176800158288744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,6144,0.13390755653381348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,5120,0.045548445648617215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,5120,0.11175555653042263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,4096,0.03701333204905192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,4096,0.09085955884721543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,10240,0.23297601275973848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,51200,0.43574221928914386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,3072,0.05803200271394518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,3584,0.079985777537028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,2560,0.05123466584417555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,3072,0.06980622477001615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,51200,1.407535129123264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,2560,0.02734044525358412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,2560,0.060164444976382785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,3584,0.07578311363855998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,2048,0.023883556326230366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,8192,0.0660053359137641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,3584,0.03347466720475091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,1536,0.01993955506218804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,1536,0.04046400056944953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,3072,0.030672887961069744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,768,0.0211217784219318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,4096,0.08107288678487141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,2048,0.04340177774429321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,512,0.01625066664483812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,768,0.025455110602908667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,1536,0.037363555696275495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,256,0.011948444777064852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,512,0.020851555797788832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,1024,0.026591999663247004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,128,0.010039110978444418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,256,0.01812088820669386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,1024,0.017642666896184284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,128,0.013819555441538492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,128,0.017551110850440133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,768,0.015847111741701763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,32,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,512,0.01475911173555586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,32,0.017276444368892245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,256,0.013609777722093793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,64,0.00904088881280687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,65536,0.4391617774963379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,65536,1.1698586146036785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,2048,0.05043377810054355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,51200,0.8335608906216092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,51200,0.35321866141425234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,1024,0.030771556827757094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,16384,0.1074133316675822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,12288,0.19841155740949842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,51200,1.158172395494249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,12288,0.0799937778049045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,12288,0.22108088599310982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,10240,0.1761866675482856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,16384,0.2751839955647786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,10240,0.06820888651741876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,8192,0.1503653393851386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,7168,0.12549955315060085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,65536,1.499431080288357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,8192,0.15150754981570774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,6144,0.10364177491929795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,7168,0.12753510475158691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,7168,0.049535112248526685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,5120,0.09276088741090562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,6144,0.11133333047231038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,4096,0.06426933076646593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,5120,0.036885334385765925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,5120,0.09355288743972778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,3584,0.05633422401216295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,8192,0.05472977624999153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,4096,0.0754151079389784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,3072,0.04785866538683573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,3584,0.06765866941875882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,64,0.017265778448846605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,2560,0.04450666573312548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,4096,0.031178666485680476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,2560,0.05084888802634346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,2048,0.033519999848471746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,3584,0.028456889920764502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,16384,0.3403617805904812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,1536,0.027657777070999146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,2048,0.04269244604640537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,3072,0.025906667113304138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,1024,0.019633778267436557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,1536,0.01739377776781718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,1536,0.03484444485770331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,1024,0.015112888481881885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,1024,0.026113778352737427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,10240,0.1793431176079644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,512,0.012058667010731168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,2560,0.023256889647907678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,768,0.014155555102560254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,768,0.02160355614291297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,256,0.009729777773221334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,512,0.013101333545313941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,128,0.008363555702898238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,6144,0.04268355502022637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,128,0.012438221938080259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,128,0.014916444818178812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,64,0.007290666302045186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,768,0.015802666544914246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,32,0.007640889121426477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,64,0.015060444672902426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,32,0.01479111115137736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,2048,0.019886222150590684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,3072,0.058964444531334766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,65536,0.4170862303839789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,51200,0.7680568695068359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,51200,0.32860178417629665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,65536,1.2752692964341905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,16384,0.2948053412967258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,256,0.012087999946541257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,51200,0.9703973134358724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,16384,0.09701066546969944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,16384,0.27628178066677517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,65536,1.031126234266493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,10240,0.17219555377960205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,12288,0.19237866666581896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,10240,0.0647066699133979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,10240,0.16198844379848903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,8192,0.051464888784620494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,7168,0.12254399723476833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,8192,0.12957421938578287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,512,0.018106665876176622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,6144,0.09624711010191177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,256,0.015511110424995422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,12288,0.19501511255900064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,5120,0.07878044578764173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,12288,0.07602133353551228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,6144,0.09958044687906902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,4096,0.06505333052741157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,5120,0.033703111939960055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,5120,0.0843893355793423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,4096,0.029205332199732464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,4096,0.068031112353007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,3584,0.02584444483121236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,3072,0.04687022169431051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,7168,0.04510133465131124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,7168,0.1165813340081109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,2560,0.040287110540601946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,3072,0.05359377794795566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,8192,0.13900444242689344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,2560,0.021184000704023574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,2048,0.029751112063725788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,2048,0.018644443816608854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,1536,0.02445777753988902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,3584,0.06082933478885227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,1536,0.016152888536453247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,6144,0.04047911034690009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,1536,0.031437334087159895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,1024,0.014165333575672574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,768,0.013820444544156393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,1024,0.023588443795839947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,512,0.011404444773991903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,768,0.01313333296113544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,768,0.020030222005314298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,256,0.009010666774378883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,512,0.012376888758606382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,512,0.016566221912701923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,128,0.008008889026112026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,3072,0.02307555576165517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,128,0.011508444117175208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,128,0.013796444568369122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,64,0.007357333269384172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,2560,0.04626489016744825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,64,0.01366311146153344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,2048,0.03899200095070733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,32,0.013813333378897773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,256,0.012124444047609964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,1024,0.017584888471497428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,65536,0.3547511100769043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,3584,0.055471999777687915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,51200,0.6502248975965712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,51200,0.2802213297949897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,16384,0.2159342235989041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,65536,1.0165706210666232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,12288,0.15169510576460096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,51200,0.8013022210862902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,16384,0.22755199008517793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,12288,0.06704177459081014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,12288,0.17401599884033203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,10240,0.05778488847944471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,65536,0.9199066162109375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,10240,0.14519555038876003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,7168,0.08724000056584676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,8192,0.1171848906411065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,16384,0.08560533656014337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,7168,0.04105688797103034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,6144,0.08214755853017171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,256,0.014398222168286642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,10240,0.13868089516957602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,5120,0.06817244158850776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,6144,0.03599111239115397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,6144,0.08977066808276707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,5120,0.03153955605294969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,5120,0.07598577605353461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,3584,0.04433688852522108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,8192,0.046000887950261436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,4096,0.06132444408204821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,3072,0.042529778348075024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,3584,0.024365334047211543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,3584,0.05329422156016032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,7168,0.1022791133986579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,2560,0.03523555397987366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,3072,0.047021332714292735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,2048,0.02942844563060337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,8192,0.0944897797372606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,2560,0.040821333726247154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,32,0.0075866662793689305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,1536,0.023137778043746948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,2048,0.017256889078352187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,4096,0.026497777965333726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,1536,0.015034667319721647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,1536,0.027657777070999146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,1024,0.013360000318951078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,1024,0.021320000290870667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,3072,0.02198666665289137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,768,0.012780444489585029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,768,0.017685333887736004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,2560,0.0195448895295461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,4096,0.05314400129848056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,256,0.00902844468752543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,512,0.011772444678677453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,512,0.014866666661368476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,256,0.011160888605647616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,1024,0.017650667164060805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,256,0.012940444052219391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,64,0.007288888924651676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,128,0.011275555524561139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,32,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,128,0.012415111064910889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,768,0.014473777678277759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,64,0.01219555570019616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,32,0.01278222186697854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,512,0.01165422217713462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,65536,0.5199511316087511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,51200,0.5098213089836968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,128,0.008294222255547842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,2048,0.03445600138770209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,51200,0.24048799938625762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,16384,0.18453421857621935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,12288,0.12077511681450738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,16384,0.1981359985139635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,12288,0.059359113375345864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,12288,0.1683128939734565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,65536,0.31006044811672634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,10240,0.052368889252344765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,8192,0.08304266797171699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,10240,0.13046222262912327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,7168,0.07069333394368489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,8192,0.039176887936062284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,65536,0.8715822431776258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,8192,0.10393244690365261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,6144,0.06568444437450834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,7168,0.09040266937679714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,6144,0.03159200151761373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,10240,0.1015662219789293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,5120,0.027666666441493567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,5120,0.06622666782803006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,4096,0.023967999551031325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,4096,0.053893334335751004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,3584,0.03932355509863959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,5120,0.05150489012400309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,3584,0.021749332547187805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,3584,0.048215111096700035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,3072,0.019670221540662978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,4096,0.03984888725810581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,3072,0.043214221795399986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,2560,0.01789422167672051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,2048,0.02477688921822442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,6144,0.07936888933181763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,3072,0.03380089004834493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,1536,0.020424000091022916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,2048,0.015929778416951496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,2560,0.029800001117918227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,1024,0.0148053334818946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,1536,0.025103110406133864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,2560,0.03704444567362467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,768,0.011776000261306763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,2048,0.03132889005872939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,768,0.016015999846988253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,512,0.009674666656388177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,1536,0.014489778214030795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,512,0.01368444412946701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,256,0.007335999773608313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,51200,0.6800906923082141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,1024,0.019275556008021038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,128,0.006335111127959357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,256,0.010552888943089379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,256,0.011990221838156382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,64,0.005643555687533484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,128,0.010383110907342698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,128,0.011132444772455426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,64,0.01111288865407308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,32,0.011397333608733283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,512,0.01108266661564509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,16384,0.07577866978115506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,65536,0.5029510921902126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,51200,0.3430195649464925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,32,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,65536,0.6990666919284396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,51200,0.21894666883680555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,51200,0.567903995513916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,16384,0.07023911343680488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,12288,0.10139466656578912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,16384,0.17279822296566436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,10240,0.0867404474152459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,12288,0.05634222428003947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,65536,0.2757013373904758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,10240,0.04879466692606608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,8192,0.07526133457819621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,10240,0.10764444536632961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,7168,0.06142222219043308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,8192,0.08656711048550075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,7168,0.035391112168629967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,7168,0.03310133351220025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,7168,0.07670222388373481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,6144,0.030208001534144085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,5120,0.04460444384151035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,768,0.011771555576059552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,16384,0.13916799757215711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,5120,0.05602044529385037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,1024,0.012549333274364471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,6144,0.05490488807360331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,4096,0.046105778879589505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,4096,0.021751999855041504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,12288,0.1292791101667616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,3072,0.027448000179396734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,3584,0.019381332728597853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,3584,0.041672888729307384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,2560,0.023149333066410486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,3072,0.017832888497246634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,8192,0.03739555676778158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,3072,0.0362097786532508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,2560,0.016698666744761996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,2560,0.031677332189348005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,2048,0.02660977840423584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,1536,0.015881778465376962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,6144,0.06643911202748616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,1536,0.02198666665289137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,1536,0.012862222062216865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,1024,0.011756444142924415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,5120,0.025994666748576697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,2048,0.019582221905390423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,1024,0.011359111302428775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,1024,0.01678311162524753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,768,0.010899555351999072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,768,0.013981334037250943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,3584,0.03209599852561951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,256,0.006314666734801398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,512,0.010337777435779572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,512,0.012098666694429187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,128,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,256,0.010062221851613786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,256,0.010624888870451184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,128,0.010084444450007545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,4096,0.035078220897250705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,128,0.009948444035318162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,64,0.009740444521109263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,512,0.007755555212497711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,2048,0.015824889143308003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,65536,0.3791333304511176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,65536,0.2275386651357015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,51200,0.32466310924953884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,16384,0.10804710785547893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,51200,0.18062044514550102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,51200,0.4488675329420302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,12288,0.08060444725884332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,16384,0.14384710788726807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,32,0.010060444474220276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,10240,0.06900178061591254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,12288,0.10598222414652507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,10240,0.04083999991416931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,65536,0.5607191191779243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,8192,0.05843200286229452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,768,0.009694221946928237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,8192,0.03150577677620782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,8192,0.07172889179653592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,7168,0.02859555681546529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,7168,0.06358133422003852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,64,0.004905777672926585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,6144,0.05535022086567349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,5120,0.03460622164938185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,6144,0.02534399926662445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,7168,0.04821955495410495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,4096,0.02663999961482154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,5120,0.046872887346479625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,12288,0.048143111997180514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,6144,0.04165510998831855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,4096,0.03859644465976291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,3584,0.017608889275126986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,10240,0.08852799733479817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,3072,0.021015110943052504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,32,0.0052391112678580815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,2560,0.01804266704453362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,3072,0.01642666591538323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,16384,0.05982222159703573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,3072,0.030457778109444514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,2048,0.015461333923869662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,2560,0.015184000134468079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,1536,0.012747555143303342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,2048,0.013903111219406128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,2048,0.022563555174403723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,1024,0.010434666441546546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,1536,0.018625777628686693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,3584,0.03421422176890903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,1024,0.01106844428512785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,768,0.008644444247086843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,768,0.010425777898894416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,5120,0.022272888157102797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,512,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,2560,0.02630666726165348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,256,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,3584,0.023718222975730896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,512,0.010668444136778513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,1536,0.012124444047609964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,256,0.009385777844323052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,256,0.00905777762333552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,64,0.00490133340160052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,1024,0.014489778214030795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,32,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,64,0.008438222110271454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,768,0.012077333198653327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,32,0.00868977771864997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,512,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,4096,0.018989332848125033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,128,0.005584888988071018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,51200,0.23655377493964302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,16384,0.07911733123991224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,65536,0.4433537854088678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,51200,0.14411377906799316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,128,0.008376888930797577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,12288,0.05791199869579739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,16384,0.049914668003718056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,16384,0.11606132984161377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,12288,0.03892533315552606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,12288,0.08678933646943833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,65536,0.2997368971506755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,8192,0.03557777735922072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,65536,0.18048532803853354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,128,0.00942311104800966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,8192,0.059154669443766274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,51200,0.3501457903120253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,10240,0.048376000589794584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,6144,0.027088888817363318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,7168,0.05213155680232578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,10240,0.03422844409942627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,10240,0.07288889090220134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,6144,0.045888000064426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,5120,0.018603555030292936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,4096,0.019582221905390423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,7168,0.02294844388961792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,4096,0.03260711166593764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,4096,0.016160888804329764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,6144,0.02091377807988061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,3072,0.015269334117571512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,3584,0.015104000767072042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,3584,0.02898400028546651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,2560,0.01347733371787601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,3072,0.014144000079896716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,3072,0.025795555777019922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,2048,0.011615999870830111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,2560,0.012938666674825879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,2560,0.02235377828280131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,1536,0.009927999642160203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,2048,0.019317333896954853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,3584,0.017808889349301655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,1024,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,1536,0.010732444624106089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,1536,0.016011555989583332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,768,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,1024,0.010279999838935005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,8192,0.025582222474945918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,7168,0.032024890184402466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,768,0.010760889285140567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,512,0.00902844468752543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,2048,0.011746666497654386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,256,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,256,0.008763555851247575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,5120,0.023730667101012334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,128,0.00461422238085005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,128,0.008683555656009251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,1024,0.012458666331238218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,64,0.004247111164861255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,128,0.007631999750932057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,64,0.007381333245171442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,512,0.009382222261693742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,65536,0.2220568921830919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,256,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,65536,0.36728088061014813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,51200,0.1835635503133138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,5120,0.03869511021508111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,16384,0.05911733044518364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,51200,0.1428097751405504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,32,0.0074480002125104266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,12288,0.043839110268486865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,16384,0.04866755670971341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,12288,0.03806489043765598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,10240,0.03886933459175958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,512,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,10240,0.0339955555068122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,10240,0.06142399708429972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,8192,0.02954933378431532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,16384,0.09603466590245564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,8192,0.025215999947653875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,7168,0.0258933338854048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,8192,0.04962933394643995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,6144,0.02327022287580702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,7168,0.02299022177855174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,65536,0.17768177721235487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,768,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,5120,0.020113777783181932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,6144,0.038464887274636164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,4096,0.017304889029926725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,5120,0.01829777823554145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,51200,0.28926489088270396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,3584,0.015263110399246216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,4096,0.016199999385409884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,4096,0.02732622292306688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,3072,0.013167999684810638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,3584,0.02492977844344245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,12288,0.07226755883958605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,2560,0.011695110963450538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,32,0.004570666700601578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,6144,0.02067377832200792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,2048,0.010081777969996134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,2560,0.01278133359220293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,2560,0.01921688848071628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,5120,0.032872888776991106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,1536,0.008773333496517604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,2048,0.016529776983790927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,1536,0.010788444843557147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,3584,0.01516266663869222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,1024,0.010060444474220276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,1024,0.010862222148312463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,3072,0.02195377813445197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,768,0.009373333719041612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,2048,0.011734222372372946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,512,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,768,0.009703999592198266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,1024,0.00693866651919153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,1536,0.013807111316257052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,256,0.005013333426581489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,512,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,256,0.00871022211180793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,7168,0.04429866539107429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,256,0.007039111521508958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,64,0.0038728887836138406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,128,0.00868444475862715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,32,0.003896000070704354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,64,0.006608888920810487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,32,0.00664444433318244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,65536,0.18998577859666613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,512,0.00906133320596483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,128,0.004566222015354368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,3072,0.014023999373118082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,65536,0.3193884425693088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,16384,0.04843644301096598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,128,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,12288,0.03545244534810384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,16384,0.03773777683575948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,16384,0.08442666795518662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,10240,0.028472890456517536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,12288,0.06410221921073066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,51200,0.14672533671061197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,768,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,8192,0.023037332627508376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,10240,0.053499556250042386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,7168,0.02049422264099121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,8192,0.019132445255915325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,8192,0.04348533352216085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,51200,0.25204266442192924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,7168,0.01772444446881612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,7168,0.03883466786808438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,5120,0.015677332878112793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,6144,0.01590666671593984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,12288,0.029706666866938274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,6144,0.034154666794670954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,4096,0.013247110777431063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,5120,0.0143857780430052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,3584,0.012166221936543783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,65536,0.1330462164349026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,4096,0.01279022213485506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,3072,0.011051555474599203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,3584,0.01202666676706738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,51200,0.10743288861380683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,3584,0.021977777282396953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,2560,0.009736888938479954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,3072,0.011349333657158745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,6144,0.018438221679793466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,2560,0.010780444575680627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,2560,0.017204445269372728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,1536,0.007229333122571309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,2048,0.010321777727868822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,2048,0.014895111322402954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,1024,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,5120,0.02922044528855218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,1536,0.012475555141766867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,4096,0.02422488894727495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,768,0.0052897776994440295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,1024,0.009856888817416297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,768,0.00867377801073922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,768,0.00868622213602066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,512,0.00833155545923445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,512,0.007641777396202087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,256,0.004137777619891697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,256,0.007976888782448238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,10240,0.02607111136118571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,128,0.00360000009338061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,1024,0.009020444419648912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,128,0.00795022232664956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,64,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,128,0.006060444646411472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,512,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,3072,0.019878221882714164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,65536,0.11687733067406549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,32,0.006234666539563074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,65536,0.13151821825239393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,65536,0.2755173312293159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,256,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,16384,0.03312622176276313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,51200,0.106240000989702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,51200,0.21628888448079428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,32,0.003528000166018804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,12288,0.023354666100607976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,16384,0.03732622332043118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,64,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,12288,0.029096888171301946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,12288,0.054936889145109385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,8192,0.016422222057978313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,51200,0.0927191111776564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,10240,0.04633422361479866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,1536,0.009718221922715506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,8192,0.01867111192809211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,8192,0.038246221012539335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,7168,0.017469333277808297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,7168,0.03438399897681342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,2048,0.008382221890820397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,16384,0.07332089212205675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,5120,0.011895111037625207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,6144,0.015837333268589444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,10240,0.025311999850802954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,5120,0.014135110709402295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,7168,0.015152000718646579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,4096,0.012679999901188744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,4096,0.021329777108298406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,3584,0.009701333112186855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,6144,0.013542222479979197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,3072,0.008541333178679148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,3584,0.011909333368142446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,10240,0.020167110694779288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,3584,0.019326221611764696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,2560,0.007689778175618913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,3072,0.011137777732478248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,2048,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,2560,0.010764444039927589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,4096,0.009860444400045607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,2560,0.015364444918102689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,1536,0.006024888820118374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,2048,0.010419555836253695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,2048,0.013464888764752282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,1536,0.009644444617960189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,1536,0.011424889167149862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,768,0.004616888860861461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,1024,0.009023110899660323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,512,0.0041751112374994485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,6144,0.02994933393266466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,3072,0.01756266587310367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,256,0.00388355553150177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,512,0.008015111088752747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,512,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,5120,0.025390222668647766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,128,0.003499555504984326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,256,0.008015111088752747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,256,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,128,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,32,0.0032017777363459268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,1024,0.00903733323017756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,64,0.005658666706747479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,65536,0.0752568907207913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,32,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,51200,0.0610391100247701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,65536,0.13129956192440456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,65536,0.23781689008076987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,16384,0.02500266664557987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,51200,0.1052631139755249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,51200,0.18568266762627494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,128,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,16384,0.036884443627463445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,16384,0.06235644552442762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,10240,0.015071999695565967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,12288,0.027985778119828966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,768,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,8192,0.013621333572599622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,10240,0.025390222668647766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,10240,0.04023289018207126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,7168,0.012127111355463663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,1024,0.004937777916590373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,8192,0.032736000087526106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,64,0.003244444520937072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,6144,0.011058666639857821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,7168,0.018630221486091614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,12288,0.016534222496880423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,5120,0.009777777724795872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,6144,0.016191111670600045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,4096,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,6144,0.025790222816997107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,5120,0.015207999282413058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,3584,0.007685333490371704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,5120,0.022427555587556627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,4096,0.018906666172875296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,3584,0.01242844429281023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,8192,0.018948444061809115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,3584,0.017363554901546903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,2560,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,3072,0.01131199962562985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,3072,0.01573244399494595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,7168,0.02957333458794488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,768,0.008664000365469191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,12288,0.04859822326236301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,4096,0.013454222016864352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,1536,0.0052426668504873914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,2048,0.01038577738735411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,3072,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,1536,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,1536,0.010399999717871347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,768,0.003937777959638172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,1024,0.008560889297061497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,2048,0.0057706667317284485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,512,0.0038951109680864546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,768,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,512,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,256,0.0032293332947625052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,1024,0.0047733332547876565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,256,0.007821333077218797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,1024,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,128,0.0031973332580592898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,2560,0.010799110763602786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,2560,0.01386666629049513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,64,0.0028737777223189673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,2048,0.012083555261294046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,32,0.0031911111954185697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,64,0.0053288886944452924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,65536,0.07360800107320149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,256,0.005662222289376789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,32,0.0052435555391841466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,51200,0.05675644344753689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,16384,0.018951111369662814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,768,0.008671110702885522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,65536,0.23694043689303926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,128,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,12288,0.01352266636159685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,512,0.006656000183688269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,16384,0.062353776560889355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,12288,0.0476746658484141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,7168,0.010055999788973067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,10240,0.040122667948404946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,6144,0.010773333410422007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,8192,0.032631999916500516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,5120,0.009380444056457942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,7168,0.029310223129060533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,4096,0.007963555554548899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,6144,0.025447999437650044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,128,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,3584,0.007637333538797166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,5120,0.02235466738541921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,10240,0.012140444583363004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,4096,0.01887999971707662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,2048,0.0052800000541739995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,3072,0.015643555257055495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,2560,0.013765333427323235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,2048,0.012111110819710625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,1024,0.004266666869322459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,1536,0.010400888820489248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,768,0.0039057777159743835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,1024,0.008376888930797577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,512,0.003583999971548716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,51200,0.1854719983206855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,768,0.007677333222495184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,256,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,128,0.003198222153716617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,512,0.006646222124497096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,1536,0.00490044429898262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,256,0.005633777628342311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,32,0.0032222221295038858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,128,0.0052764444715446895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,65536,0.07121777534484863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,32,0.004938666605287128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,3072,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,51200,0.05684266487757365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,16384,0.016203555795881484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,8192,0.010732444624106089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,12288,0.011402666568756104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,51200,0.19008444415198433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,16384,0.06191644403669569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,10240,0.010729778144094678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,8192,0.009637333452701569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,12288,0.04764444298214383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,7168,0.009372444616423713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,10240,0.039748443497551814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,6144,0.008343111309740279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,2560,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,7168,0.029056890143288508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,3584,0.017214222086800467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,4096,0.009673777553770278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,64,0.0028897778441508612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,5120,0.022268444299697876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,3584,0.00902844468752543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,3072,0.008024888734022776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,4096,0.018777777751286823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,2560,0.007294221884674496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,8192,0.032223112053341336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,5120,0.011394666300879585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,2560,0.013796444568369122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,1536,0.004622222234805425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,2048,0.012159110771285163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,1536,0.010394666757848527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,64,0.0052791109515561005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,768,0.0035422220826148987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,1024,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,3072,0.0155395550860299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,2048,0.005305777821275923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,65536,0.23757955763075086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,256,0.003208000005947219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,6144,0.02558044426971012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,128,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,64,0.002532444480392668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,128,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,32,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,3584,0.017289777596791584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,32,0.004938666605287128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,768,0.007675555845101674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,512,0.006660444455014334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,256,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,1024,0.003933333274390963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,64,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,512,0.003256888853179084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,16384,3.676881578233507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,16384,4.420775943332248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,16384,1.1998426649305556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,51200,4.633357153998481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,16384,3.8050765991210938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,12288,2.886964374118381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,12288,0.9145875506930881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,12288,3.1195066240098743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,10240,2.9102630615234375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,10240,0.7759840223524305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,10240,2.5221689012315536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,12288,2.751424789428711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,10240,2.331648084852431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,51200,12.001135084364149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,8192,2.024526172214084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,8192,0.6030862066480849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,8192,1.9875058068169489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,8192,1.86499998304579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,7168,0.5333609051174587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,7168,1.749864896138509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,7168,1.682588471306695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,6144,0.47732798258463544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,6144,1.7059173583984375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,7168,1.5977795918782551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,6144,1.4316737916734483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,5120,1.1941875881618924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,5120,0.4120791223314073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,6144,1.3669101927015517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,4096,0.8767040040757922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,5120,1.2700604332817924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,4096,0.33634577857123477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,5120,1.1478658252292209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,4096,0.9615777333577474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,3584,0.7523671256171333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,3584,0.2962648868560791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,4096,0.9203555848863391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,3072,0.6857244703504773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,3584,0.789954662322998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,3072,0.7255715794033475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,3072,0.6789280043707954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,3584,0.8236986796061198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,51200,12.652254740397135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,51200,18.59156968858507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,2560,0.6055306858486599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,2048,0.4643591245015462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,2560,0.5611324310302734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,2048,0.19704622692532012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,1536,0.34560712178548175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,2048,0.5349271032545302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,2560,0.6550969017876519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,1024,0.23719912105136445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,2560,0.23016977310180664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,1536,0.36319377687242294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,3072,0.271783987681071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,768,0.19601955678727892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,1024,0.2697510984208849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,2048,0.4751822153727214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,768,0.22068799866570366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,768,0.10924710830052693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,1024,0.29179289605882436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,512,0.1983546680874295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,512,0.18009511629740396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,256,0.1058880024486118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,768,0.23207378387451172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,256,0.1783813370598687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,1536,0.16957778400845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,128,0.08138400316238403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,1024,0.1247840060128106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,128,0.10906044642130534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,512,0.0975946651564704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,64,0.07003733184602526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,128,0.08453510867224799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,32,0.07140355639987521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,512,0.16474221812354192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,1536,0.3987270991007487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,32,0.17411555184258354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,128,0.16793066925472686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,256,0.14275733629862467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,256,0.08637422323226929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,64,0.17124888632032606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,65536,4.781212276882596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,51200,3.7130567762586804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,16384,3.520079082912869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,51200,9.706061469184027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,65536,12.041827731662325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,16384,0.9520542356703016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,16384,3.678345574273003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,16384,2.976590262518989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,12288,2.5387467278374567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,51200,13.890315585666231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,12288,0.6863795386420356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,12288,2.3970247904459634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,12288,2.1826231214735246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,10240,0.5883324411180284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,10240,2.08809683057997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,51200,10.146095275878906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,10240,2.03147824605306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,8192,1.955118179321289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,8192,0.5358364317152235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,65536,18.476664225260418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,10240,1.7918559180365667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,8192,1.5413440068562825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,7168,1.3557013405693903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,65536,14.041730244954428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,8192,1.4380559921264648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,7168,1.3779137929280598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,6144,1.0901048448350694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,6144,0.37543288866678876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,6144,1.1125208536783855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,7168,0.4217200014326308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,5120,0.99526670244005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,7168,1.2241217295328777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,5120,0.3214648829566108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,6144,1.0269483990139432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,4096,0.2643866539001465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,4096,0.841759999593099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,5120,0.8471022182040744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,4096,0.762864907582601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,4096,0.6762586699591743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,3584,0.23344622717963326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,3072,0.5317102008395725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,3072,0.20843911170959473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,3584,0.6207795672946507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,3072,0.5572471088833278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,3072,0.5305617650349935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,3584,0.6356222364637587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,2560,0.18944977389441597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,2560,0.4711333380805121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,2048,0.3865288893381755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,2560,0.45246844821506077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,2048,0.15562310483720568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,1536,0.30525954564412433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,3584,0.6499182383219401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,2048,0.37673245535956484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,1536,0.14156444867451987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,2560,0.4992435773213704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,1536,0.31251200040181476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,1536,0.2844533390469021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,768,0.1506106721030341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,1024,0.25269420941670734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,1024,0.216285334693061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,5120,0.936896006266276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,768,0.18313954936133492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,2048,0.3862542311350505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,768,0.1748755507998996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,1024,0.18841155370076498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,256,0.08688000175688003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,512,0.1566728883319431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,512,0.1425982183880276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,128,0.06603733036253187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,1024,0.10080977943208481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,256,0.11307911078135173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,256,0.14089155197143555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,512,0.11847378147972955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,128,0.0680382251739502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,128,0.0872560011015998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,32,0.06342311037911309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,512,0.07899644639756945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,64,0.1351626714070638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,32,0.13752888308631048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,256,0.07032977872424655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,64,0.05956355730692545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,768,0.08930755323833889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,65536,1.4441564348008897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,128,0.13257955180274114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,51200,3.187520980834961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,65536,3.9413405524359812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,51200,1.094564437866211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,16384,0.4019351005554199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,16384,1.142823961046007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,16384,1.0681991577148438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,16384,0.9297973844740125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,51200,4.448368920220269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,12288,0.8819182184007434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,65536,5.694352891710069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,12288,0.2546159956190321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,65536,4.080659654405382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,12288,1.004808849758572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,10240,0.6283466551038954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,12288,0.6604977713690864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,10240,0.20293600029415557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,8192,0.5216266844007704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,10240,0.7068595356411405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,10240,0.5354488690694174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,51200,3.208030276828342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,8192,0.5026346842447916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,8192,0.48472266727023655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,6144,0.3953919940524631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,7168,0.470536020067003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,7168,0.3863244321611192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,6144,0.13162310918172201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,5120,0.30344266361660427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,6144,0.3253377808464898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,7168,0.41679022047254777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,5120,0.30732621086968315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,4096,0.22492265701293945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,7168,0.14453066719902888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,4096,0.09002488851547241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,3584,0.20082399580213758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,6144,0.3738551139831543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,8192,0.1603413290447659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,3584,0.21460355652703178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,3584,0.19774756166670057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,3072,0.17464533117082384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,5120,0.28321777449713814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,4096,0.24640978707207573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,3072,0.07491466734144422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,2560,0.15064533551534018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,2560,0.15966755814022487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,5120,0.11143910884857178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,2560,0.15144266022576228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,2048,0.13421866628858778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,3072,0.18848089377085367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,2048,0.1323466698328654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,3072,0.1741928921805488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,2048,0.05474844574928284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,1536,0.04541688826349047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,1536,0.10497155454423691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,1024,0.06604888704088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,1536,0.10241422388288711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,2560,0.06565955612394545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,768,0.05263911022080315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,1024,0.03846222162246704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,1024,0.07796977625952826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,4096,0.22967823346455893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,512,0.04504177636570401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,768,0.06433333290947808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,768,0.06520266665352716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,512,0.055267555846108325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,256,0.02790844440460205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,512,0.05350044369697571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,1024,0.0800079968240526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,256,0.04941599898868137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,256,0.0421688887808058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,256,0.029112001260121662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,2048,0.12374399767981635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,768,0.03533777925703261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,64,0.0207013338804245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,128,0.028365333875020344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,128,0.046776887443330556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,32,0.021399110555648804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,128,0.034612443712022566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,64,0.0472942226462894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,32,0.048044443130493164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,128,0.02422399984465705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,1536,0.08492355876498753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,512,0.03233333428700765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,65536,1.0861049228244357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,3584,0.08115377691056994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,65536,2.8834711710611978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,51200,2.3876027001274958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,51200,0.8736640082465278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,16384,0.754492441813151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,16384,0.3120177851782905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,16384,0.8226542472839355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,12288,0.5635831091139052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,51200,3.3913813696967234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,16384,0.6976159943474664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,12288,0.18018666903177896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,65536,3.1123964521620007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,12288,0.5708355373806423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,10240,0.43174489339192706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,51200,2.4485840267605252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,12288,0.4903048939175076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,10240,0.17089333799150255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,10240,0.48905335532294375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,8192,0.12511555353800455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,10240,0.4052080048455133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,7168,0.3417750994364421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,65536,4.294295840793185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,8192,0.35802398787604434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,7168,0.11187556054857041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,6144,0.27656179004245335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,7168,0.290321773952908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,6144,0.09836533334520127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,6144,0.28448443942599827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,6144,0.25188978513081867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,8192,0.4002186722225613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,5120,0.084370666080051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,4096,0.18350399865044487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,5120,0.23661777708265516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,5120,0.2117840051651001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,3584,0.1539751158820258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,4096,0.1867306629816691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,7168,0.32949778768751353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,4096,0.1724231110678779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,3072,0.13158310784233943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,3584,0.16465511586931017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,5120,0.2542640103234185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,2560,0.11701244778103298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,3072,0.144923554526435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,3072,0.135970671971639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,8192,0.42157332102457684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,2560,0.12377777364518906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,2048,0.0968773365020752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,3584,0.06542488601472643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,2048,0.10261778036753337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,1536,0.06734577814737956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,2048,0.04424800144301521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,3072,0.05652088589138455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,1536,0.08176977766884698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,2560,0.05099199877844917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,1024,0.049880888726976186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,2560,0.11686578061845566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,1024,0.0623555580774943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,2048,0.09830400016572739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,3584,0.15409333176083034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,768,0.03995733459790548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,1024,0.030963556634055242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,768,0.028431110911899145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,768,0.050251556767357715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,768,0.051333334710862905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,1536,0.036293334431118436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,512,0.02619022296534644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,256,0.02212000058756934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,512,0.04282133446799385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,1024,0.06201777855555216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,256,0.02310488952530755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,256,0.03813777698410882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,128,0.019681778219011094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,256,0.03403910994529724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,64,0.016695110334290397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,128,0.022672888305452134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,128,0.0359919998380873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,128,0.028408888313505385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,32,0.03674044542842441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,4096,0.07056978013780382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,32,0.016922665966881644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,64,0.03654311100641886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,1536,0.08148622512817383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,65536,0.9218880335489908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,65536,2.4868782891167536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,512,0.04286844531695048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,51200,1.9503947363959417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,16384,0.6976408958435059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,51200,0.7044257587856717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,65536,3.897674560546875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,16384,0.2473857667711046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,16384,0.693453311920166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,16384,0.5654888682895237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,65536,2.5861600240071616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,512,0.030044443077511255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,12288,0.15506311257680258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,12288,0.5182551278008355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,10240,0.41839199595981175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,51200,2.9892639584011502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,10240,0.1324924495485094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,10240,0.3959662119547526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,8192,0.3152035607231988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,10240,0.3414444393581814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,8192,0.10992622375488281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,8192,0.3140159977806939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,12288,0.5311013327704536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,8192,0.2749226623111301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,7168,0.09461600250667995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,7168,0.27833154466417104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,12288,0.4188008838229709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,7168,0.24745599428812662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,6144,0.08436977863311768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,6144,0.2404471238454183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,5120,0.17979378170437285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,6144,0.20927822589874268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,4096,0.13833955923716226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,5120,0.19898577531178793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,7168,0.29762665430704754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,5120,0.17767466439141166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,3584,0.14310044712490505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,6144,0.2616364426083035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,4096,0.1486497720082601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,3584,0.14043466250101724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,3072,0.10335555341508652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,51200,2.0126586490207248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,3584,0.13165067301856145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,3072,0.04810666706826952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,5120,0.07178399960199992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,3072,0.11618844668070476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,4096,0.06151466899447971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,2560,0.10529955228169759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,2048,0.07772266864776611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,3584,0.05565422111087375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,2048,0.08744888835483128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,1536,0.057131555345323354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,2048,0.08357155323028564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,2560,0.09961422284444173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,3072,0.12268800205654568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,1024,0.045250667466057666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,1536,0.031722668144438006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,1536,0.06951555278566149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,2560,0.04357599880960253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,1536,0.06803288724687365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,768,0.033181332879596286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,1024,0.052928000688552856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,1024,0.05382666654056973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,768,0.023746665981080797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,768,0.04206044475237528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,768,0.04430222180154589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,2048,0.03795555565092299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,256,0.019701333509551156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,512,0.035232888327704534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,512,0.03682311044798957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,4096,0.15808978345659044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,256,0.030989332331551447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,2560,0.09938133425182766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,256,0.01937599976857503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,256,0.02906844351026747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,512,0.02565333412753211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,128,0.019870221614837646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,128,0.029630220598644678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,32,0.01719999975628323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,512,0.021427555216683283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,64,0.029341333442264136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,32,0.02938844429122077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,1024,0.025053333905008104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,64,0.016622222132152982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,128,0.024783111280865137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,65536,0.7601146697998047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,128,0.016895999511082966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,65536,1.9773794809977214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,16384,0.5082240104675293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,51200,1.5823947058783636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,16384,0.5552755461798774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,16384,0.16855288876427543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,65536,3.053036371866862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,12288,0.3338328997294108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,65536,2.1310062408447266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,51200,2.406806310017904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,12288,0.38813334041171604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,10240,0.33838844299316406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,12288,0.3504062228732639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,10240,0.11662133534749348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,10240,0.3949688805474176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,8192,0.26852620972527397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,10240,0.2861413425869412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,16384,0.47323020299275714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,7168,0.23260890112982857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,8192,0.2661120096842448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,8192,0.23257954915364584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,51200,0.614325311448839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,6144,0.19724622037675646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,7168,0.20405777295430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,51200,1.692238171895345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,6144,0.19468622737460664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,5120,0.17203466097513834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,6144,0.17699289321899414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,5120,0.06021866533491346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,7168,0.07900800307591756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,4096,0.13353155718909368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,7168,0.2269653214348687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,4096,0.050155556864208646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,4096,0.13125422265794542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,6144,0.06999733050664265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,3584,0.11975910928514268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,8192,0.08976888656616211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,3584,0.04507466819551256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,3072,0.10159999794430202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,3584,0.11645778020222981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,5120,0.16318310631646052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,3584,0.11124177773793538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,2560,0.07950488726298015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,3072,0.040842665566338435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,3072,0.10115466515223186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,2048,0.06853777832455106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,2560,0.08717511097590129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,2560,0.0845937795109219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,4096,0.1257715490129259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,2048,0.07215911149978638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,2048,0.07146666447321574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,12288,0.13109689288669163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,1536,0.057413332992129855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,1536,0.025957332717047796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,1536,0.05864533450868395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,3072,0.09771200021107991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,1024,0.021697777840826247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,1536,0.04895377821392483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,768,0.0358551111486223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,1024,0.03538577755292257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,2560,0.036751998795403376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,768,0.01956888867749108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,768,0.038718223571777344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,1024,0.045886221859190196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,768,0.02574933403068119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,256,0.015800888339678448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,512,0.01824711097611321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,512,0.03222399950027466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,5120,0.1499333381652832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,256,0.017049777838918898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,128,0.013263111313184103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,256,0.025366221865018208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,512,0.020259555843141343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,128,0.016578666037983365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,128,0.02475022276242574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,32,0.012614222036467658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,128,0.021446221404605444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,1024,0.043954667117860585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,2048,0.03090133269627889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,64,0.024471999870406255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,32,0.024570667081409033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,256,0.025664000047577754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,64,0.012146666646003723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,65536,0.644120004442003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,512,0.02932000160217285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,51200,1.645133336385091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,65536,1.9784213172064886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,51200,0.518855094909668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,16384,0.5377146402994791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,65536,1.9611360761854384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,16384,0.4969200028313531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,16384,0.14894666936662462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,51200,1.923144022623698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,12288,0.3928835656907823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,12288,0.11502667268117268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,12288,0.34931643803914386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,10240,0.3008808824751112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,12288,0.31049956215752494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,65536,2.5635768042670355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,10240,0.11371821827358669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,8192,0.22727110650804308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,10240,0.2918817732069227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,10240,0.25961245430840385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,8192,0.22755556636386445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,8192,0.2115831110212538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,7168,0.07121421893437703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,7168,0.20234666930304634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,7168,0.1853004429075453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,6144,0.17304355568355986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,6144,0.16159021854400635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,5120,0.13933244016435412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,16384,0.43257954385545516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,5120,0.05415111117892795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,6144,0.1740657753414578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,5120,0.13931200239393446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,4096,0.11700356006622314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,4096,0.11584533585442437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,4096,0.04528977804713779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,3584,0.09683822260962592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,5120,0.1463573376337687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,3584,0.10386311345630223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,4096,0.11766221788194443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,3584,0.10335199700461493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,3584,0.04087022278043959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,3072,0.036814222733179726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,3072,0.0907813310623169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,2560,0.06222133504019844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,8192,0.08165155516730414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,2560,0.07798844575881958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,2560,0.03266933229234483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,2048,0.05381155676311917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,2560,0.07800088988410102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,51200,1.4563271204630535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,1536,0.0473306675752004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,3072,0.0807173318333096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,2048,0.06465155548519559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,2048,0.06554666492674086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,6144,0.06262311008241442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,1024,0.030001779397328694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,1536,0.05192977852291531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,3072,0.09194844298892552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,768,0.02349511120054457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,7168,0.19627732700771758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,1024,0.020460445019933913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,1024,0.03939377930429246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,512,0.01752622259987725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,768,0.018445332845052082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,768,0.032207998964521624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,768,0.03492888808250427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,256,0.01885333326127794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,2048,0.028620445066028174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,512,0.030052443345387776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,1536,0.023670223024156358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,256,0.023111111587948267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,128,0.016286222471131217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,256,0.023186667097939387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,1536,0.05312088794178433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,64,0.015166223049163818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,128,0.01605244477589925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,1024,0.04193511274125841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,128,0.019468444916937087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,64,0.022006221943431433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,512,0.016859556237856548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,256,0.015895111693276297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,128,0.022110222114457026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,32,0.015536889433860779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,512,0.026018665896521673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,65536,1.7278515497843425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,51200,1.2568800184461806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,65536,1.605324427286784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,65536,0.616321775648329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,16384,0.42322667439778644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,51200,1.660190264383952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,16384,0.13830488257937962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,51200,1.2046675152248807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,16384,0.3540808889600966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,12288,0.2980399926503499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,12288,0.26922665701972115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,32,0.021695110532972548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,10240,0.24096533987257215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,16384,0.43163466453552246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,12288,0.2664079931047228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,10240,0.2189075549443563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,65536,2.18178579542372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,12288,0.10365688800811768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,8192,0.19996266894870332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,8192,0.07077599896325006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,8192,0.17998133765326607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,10240,0.0897128913137648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,10240,0.2543315622541639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,7168,0.06449688805474176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,51200,0.49741331736246747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,7168,0.15865066316392687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,6144,0.056925336519877114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,6144,0.15117243925730386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,6144,0.13822221755981445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,5120,0.1181288825141059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,5120,0.0488666660255856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,5120,0.12722222010294595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,4096,0.09616355763541327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,5120,0.11754755179087321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,7168,0.17541333039601645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,8192,0.19975554943084717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,3584,0.07716888851589628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,4096,0.040200889110565186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,4096,0.1017964416080051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,3072,0.07570044199625652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,4096,0.09757688972685073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,3584,0.09141955773035686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,7168,0.1660711103015476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,3072,0.07963111003239949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,3072,0.0770897799068027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,2560,0.06890133354398939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,2048,0.04196977615356445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,3584,0.035679999325010515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,3584,0.08851999706692165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,3072,0.03233244352870517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,1536,0.03972888986269633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,2048,0.025062221619817946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,2560,0.02914666798379686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,6144,0.14070133368174234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,1536,0.045838223563300244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,1024,0.02614044480853611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,1536,0.047057777643203735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,2048,0.05730755461586846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,1024,0.034854223330815635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,768,0.020991111795107525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,2560,0.05677422549989489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,768,0.016552888684802584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,768,0.028455999162462022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,512,0.015826667348543804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,768,0.030851556195153132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,2560,0.06666577524609037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,512,0.015607110328144498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,512,0.025807110799683466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,1024,0.018390221728218924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,1024,0.036723554134368896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,128,0.011045333411958484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,2048,0.056836446126302086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,128,0.01960355540116628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,256,0.011818666425016193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,512,0.023190221852726404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,64,0.009386666946940953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,32,0.00960888879166709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,256,0.02072622213098738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,64,0.01960000064637926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,256,0.014149333039919535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,32,0.019299555155966017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,1536,0.02160088883505927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,128,0.014294221997261047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,65536,0.5518639882405599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,65536,1.3120959599812825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,256,0.02001422146956126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,51200,0.9038107130262586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,16384,0.332352876663208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,65536,1.7802683512369792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,51200,0.4379146628909641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,65536,1.366451581319173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,128,0.017775999175177682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,16384,0.12368888325161403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,12288,0.2634284496307373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,16384,0.35440532366434735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,16384,0.31073421902126735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,10240,0.22081955273946127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,12288,0.2712373203701443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,12288,0.2334195507897271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,10240,0.2148604393005371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,8192,0.18418578306833902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,10240,0.19404533174302843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,8192,0.06685244374805026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,8192,0.17085333665211996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,7168,0.14917333920796713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,8192,0.15750133991241455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,7168,0.056008888615502246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,6144,0.11437688933478461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,12288,0.09383733405007257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,51200,1.0609110726250541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,7168,0.14044088787502712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,5120,0.1142773363325331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,10240,0.0793715582953559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,6144,0.13003644678327772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,6144,0.1223306655883789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,5120,0.04193066557248434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,5120,0.10421244303385417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,4096,0.08800088697009617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,51200,1.529024018181695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,7168,0.15438932842678493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,3584,0.0719555550151401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,6144,0.04943199952443441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,3584,0.07828088601430257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,3584,0.03163733416133457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,5120,0.10962133275138007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,3072,0.06879288620418973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,3072,0.06916533576117621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,3072,0.028343998723559912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,4096,0.034845332304636635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,2560,0.025900445050663416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,2560,0.05915821923149956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,2048,0.04317955507172478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,4096,0.08556533522076076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,3072,0.06042933464050293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,1536,0.0347377790345086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,2048,0.02253066665596432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,2048,0.04933333396911621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,2048,0.05061155557632446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,1536,0.018812444474962022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,2560,0.054029332266913525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,1536,0.04268355502022637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,1024,0.01648977730009291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,768,0.01936533384852939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,1024,0.0332293344868554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,3584,0.07834311326344807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,768,0.015218666858143277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,768,0.02513155506716834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,512,0.01648622254530589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,768,0.027793778313530818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,1024,0.025166221790843542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,256,0.012855999999576144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,512,0.0140328887436125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,4096,0.08636622296439277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,512,0.02312177750799391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,1024,0.030351112286249798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,256,0.018180444836616516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,256,0.01851733360025618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,128,0.01722755531469981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,64,0.009008000294367472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,1536,0.04018222292264303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,32,0.009330666727489894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,512,0.020609777834680345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,256,0.013239999612172445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,32,0.016911110944218107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,2560,0.05987111065122816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,128,0.01311288856797748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,128,0.016878222425778706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,65536,0.4336862299177382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,64,0.01720355616675483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,128,0.010879110958841113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,51200,0.9021502600775825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,65536,1.4536497328016493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,65536,1.171426667107476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,51200,1.1896257400512695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,16384,0.3619982136620416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,65536,1.1207093132866752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,12288,0.2052391105228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,16384,0.27853600184122723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,10240,0.17147821850246855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,12288,0.21703022056155732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,12288,0.20305866665310332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,51200,0.33483023113674587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,8192,0.14332621627383763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,10240,0.17872356043921578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,10240,0.17107021808624268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,16384,0.10168000062306722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,51200,0.8933902316623263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,7168,0.12059822347429063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,8192,0.13895732826656765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,12288,0.07866844203737047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,7168,0.047911110851499766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,16384,0.2693928877512614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,7168,0.12370222144656712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,10240,0.06707555717892118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,6144,0.04202666547563341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,5120,0.08549422025680542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,6144,0.10803111394246419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,8192,0.054104887776904635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,8192,0.1436728901333279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,4096,0.06804000006781684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,5120,0.09386755360497369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,6144,0.10535644160376655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,4096,0.07396977477603488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,4096,0.031026668018764917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,4096,0.0763911141289605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,6144,0.10858133104112412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,3072,0.04827733172310723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,3584,0.07010044654210408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,5120,0.036031110419167414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,3072,0.057745774586995445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,3072,0.06235288911395603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,3584,0.055811557504865855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,3584,0.06592533323499891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,2560,0.05286666750907898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,2048,0.03445777628156874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,5120,0.09154488643010457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,2560,0.04002755549218919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,2048,0.04229244589805603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,2048,0.019920888874265883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,2560,0.05078044533729553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,2560,0.022709333234363135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,1536,0.034020443757375084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,1536,0.017603556315104168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,1536,0.03705599904060364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,1024,0.02534488836924235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,1024,0.029040889607535467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,7168,0.12623288896348742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,1024,0.015173332558737861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,1024,0.019702222612169053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,768,0.014202665951516895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,768,0.021366222037209406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,2048,0.04489422175619337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,512,0.017717333303557504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,512,0.020669332808918424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,256,0.010799110763602786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,3072,0.025109334124459162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,768,0.015841777125994366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,256,0.012127999630239276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,512,0.012591111163298288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,256,0.017607111069891188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,768,0.024484443995687697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,128,0.012446222205956778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,128,0.014783999986118741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,128,0.016340444485346477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,32,0.010297777752081553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,64,0.014845333165592618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,32,0.014487110906177096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,256,0.015364444918102689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,128,0.01036533299419615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,3584,0.02776622275511424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,64,0.009977777798970541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,65536,0.4008968936072455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,512,0.013023110727469126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,51200,0.8483982086181641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,51200,0.31414222717285156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,16384,0.2693671120537652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,65536,1.01418579949273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,51200,0.9660960303412544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,16384,0.2576933436923557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,65536,1.0419884787665474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,12288,0.18348177274068198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,16384,0.23795199394226074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,12288,0.19042489263746473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,12288,0.17898221810658774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,65536,1.2636986838446724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,10240,0.06125599808163113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,10240,0.16004889541202122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,8192,0.12789155377282038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,8192,0.12827111615075007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,8192,0.12268977695041233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,12288,0.07236266798443265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,1536,0.026576000783178542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,7168,0.10700088739395142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,16384,0.10200355450312297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,7168,0.11406577958001031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,6144,0.09870666927761501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,6144,0.09837244616614448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,6144,0.09503289063771565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,10240,0.16124355792999268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,5120,0.07835911379920112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,7168,0.04221955604023404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,5120,0.0825164450539483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,5120,0.08089600006739299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,5120,0.03233866559134589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,4096,0.06690133280224271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,4096,0.06804533137215509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,8192,0.047593779034084745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,3584,0.054126222928365074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,3584,0.0601137810283237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,51200,0.7605422337849935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,3072,0.047001779079437256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,3072,0.05316444569163852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,6144,0.03719200028313531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,4096,0.02725155486000909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,2560,0.04047288828425937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,3072,0.02254933284388648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,3072,0.053957333167394005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,10240,0.152782228257921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,2560,0.046078221665488354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,2560,0.020656888683636982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,2560,0.047172443734274976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,2048,0.018177777528762817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,2048,0.03853777713245816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,1536,0.02796088986926609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,2048,0.040411555104785495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,1536,0.016023111012246873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,1536,0.031077331966824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,1024,0.019498666127522785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,1536,0.03347288899951511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,7168,0.10992088582780625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,1024,0.014004444082578024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,1024,0.025960889127519395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,768,0.019658666517999437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,768,0.02237333357334137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,512,0.01420088940196567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,3584,0.06119733386569553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,2048,0.034501334031422935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,512,0.016591111818949383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,512,0.018645332919226754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,512,0.012147555748621622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,768,0.013038222160604266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,256,0.014191110928853353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,4096,0.061405334207746715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,256,0.01629244453377194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,256,0.011688888900809817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,128,0.01348977784315745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,128,0.015495111544926962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,64,0.013449778159459433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,32,0.009351111120647853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,256,0.00905511114332411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,768,0.017307556337780423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,65536,0.7196702427334256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,128,0.011401777466138205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,65536,0.3175022337171766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,3584,0.024894222617149353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,65536,0.8218604193793403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,51200,0.59933869043986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,64,0.007061333292060428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,16384,0.18419733312394884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,51200,0.2878053453233507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,51200,0.6340071360270182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,16384,0.2321093347337511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,16384,0.08038133382797241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,12288,0.1342782179514567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,16384,0.2087848848766751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,32,0.013519110778967539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,12288,0.06079910861121284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,10240,0.11932355827755398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,12288,0.17211822668711343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,128,0.007978666987684038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,10240,0.1275635560353597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,8192,0.10023733642366196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,8192,0.11492978201972114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,8192,0.10517777999242146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,8192,0.04076711005634732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,7168,0.08686044481065537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,12288,0.15014578236473933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,1024,0.02331111166212294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,7168,0.10196089082294041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,6144,0.06987288925382826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,6144,0.08781599998474121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,6144,0.0817306637763977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,6144,0.0323884454038408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,65536,1.0494693120320637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,5120,0.07417600022421943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,7168,0.09377777576446533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,4096,0.04867644442452324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,5120,0.07104621993170844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,4096,0.06007644202974108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,5120,0.06066933605406019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,3584,0.043097777499092944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,51200,0.8427635298834907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,3584,0.05362311005592346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,5120,0.028582221931881372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,3072,0.034824887911478676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,3584,0.022134222918086584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,3072,0.0474639998541938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,3072,0.04733599887953865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,3072,0.020218667056825426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,10240,0.14247911506228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,2560,0.041197333070966936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,2560,0.04101422097947862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,3584,0.05341688791910807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,2048,0.024088889360427856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,2048,0.034577776988347374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,4096,0.05994488795598348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,2048,0.035535110367669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,1536,0.020249777370029025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,7168,0.03647822141647339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,1536,0.0147706667582194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,1536,0.028176890479193792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,1536,0.0296106669637892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,2560,0.018443556295500863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,1024,0.021477333373493616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,768,0.012231999801264869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,1024,0.023407111565272015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,4096,0.024262222978803847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,2560,0.029468443658616807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,10240,0.05299289027849833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,512,0.009486222432719337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,768,0.019663110375404358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,1024,0.014856888188256158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,256,0.007628444168302748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,512,0.0166657782263226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,1024,0.0129457778400845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,256,0.014871999621391296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,256,0.01072622256146537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,2048,0.016470222009552848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,768,0.012118221984969245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,128,0.010913777682516309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,64,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,512,0.014930667148696052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,128,0.014389332797792224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,32,0.006545777950021956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,256,0.012808889150619507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,64,0.012264000044928657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,128,0.0069751110341813825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,768,0.01776711146036784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,65536,0.6124977535671657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,128,0.012443555725945367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,512,0.011336888704035016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,65536,0.29643821716308594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,51200,0.533824019961887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,65536,0.8482044537862142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,51200,0.23292265997992623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,32,0.012335999972290464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,51200,0.6842640240987142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,16384,0.07445955276489258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,12288,0.12602133221096462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,16384,0.2001964383655124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,12288,0.14800088935428196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,12288,0.05760444535149468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,12288,0.1512382162941827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,10240,0.11244444052378337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,65536,0.7873751322428385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,10240,0.050138665570153125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,10240,0.12444267008039687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,8192,0.0859066645304362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,10240,0.12559555636511907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,7168,0.07650844256083171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,16384,0.20120978355407715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,8192,0.10031021965874566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,7168,0.03472177849875556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,6144,0.06773777802785237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,7168,0.09151733583874172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,6144,0.07776177591747709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,16384,0.16117421785990396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,8192,0.03885866536034478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,5120,0.06521155436833699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,8192,0.10314310921563043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,4096,0.044088890155156456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,5120,0.02781955401102702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,5120,0.06855466630723742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,6144,0.031288888719346784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,4096,0.023197333017985027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,5120,0.059903999169667564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,4096,0.05885066588719686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,51200,0.6041733423868815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,3072,0.03414577907986111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,3584,0.05183822247717115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,7168,0.08945600191752116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,3072,0.041386667225095965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,3072,0.01937333411640591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,2560,0.029891557163662378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,4096,0.052384889788097806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,6144,0.0797591143184238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,2048,0.02475555572244856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,2560,0.017887999614079792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,2560,0.04012444284227159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,3584,0.02167466613981459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,1536,0.020278222031063504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,2048,0.016201777590645682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,2048,0.03475555446412828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,1536,0.024732444021436904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,3072,0.046244445774290294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,3584,0.03939200109905667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,2560,0.03570488757557339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,1024,0.018458666072951425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,768,0.011902222202883827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,1024,0.012484444512261285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,1024,0.022408889399634466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,2048,0.030076444149017334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,512,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,768,0.011771555576059552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,768,0.0192586663696501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,1024,0.014929778046078153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,512,0.013270222478442721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,512,0.011049778097205691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,512,0.016562667157914903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,256,0.011728888584507836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,3584,0.04639377858903673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,128,0.006609777609507243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,128,0.01090222183201048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,1536,0.014312000738249885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,64,0.006614222294754452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,1536,0.028679112593332928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,256,0.007844444778230455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,64,0.01088444474670622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,256,0.010420444111029306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,32,0.011056889262464313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,256,0.014087110757827759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,65536,0.45460886425442165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,768,0.015606222881211175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,65536,0.26574399736192494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,128,0.013526221944226159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,128,0.01038044442733129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,51200,0.3659528891245524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,65536,0.7085484398735894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,16384,0.1315973334842258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,51200,0.21028978294796416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,65536,0.6278515391879612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,12288,0.10166666905085246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,16384,0.1784720023473104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,51200,0.4898177782694499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,16384,0.16196978092193604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,10240,0.09118133121066624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,12288,0.05445333321889242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,12288,0.12210310830010308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,10240,0.04582133226924472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,10240,0.10933599869410197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,10240,0.10399733649359809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,16384,0.06893155309889051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,8192,0.08723644415537517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,8192,0.03390311201413473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,8192,0.08577333556281196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,12288,0.13178755177391901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,7168,0.03183199961980184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,32,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,8192,0.07613244321611193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,51200,0.5543422169155544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,6144,0.06597866614659627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,5120,0.04337777694066366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,6144,0.028011557128694322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,5120,0.055529776546690196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,5120,0.024116444918844435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,4096,0.03389777739842733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,7168,0.07656888829337226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,6144,0.05339644352595011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,4096,0.02067466742462582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,4096,0.04566311173968845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,4096,0.04893866512510511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,7168,0.06490577591790093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,3584,0.04125688804520501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,3072,0.026744888888465032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,3584,0.043748444981045194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,7168,0.07605066564348009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,5120,0.05783200263977051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,2560,0.022825777530670166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,3072,0.039084444443384804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,3584,0.030619555049472388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,2560,0.016198222835858662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,2048,0.01947111056910621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,2560,0.03341777788268195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,6144,0.06646488772498237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,2048,0.014532445205582513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,2048,0.029375112719006006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,3072,0.01715377800994449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,3072,0.03654133280118307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,1024,0.010948444406191507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,1536,0.01278133359220293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,2560,0.031599998474121094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,1024,0.016712889075279236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,1024,0.011407111254003314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,3584,0.01869155632125007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,2048,0.02660266723897722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,768,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,768,0.014138667119873894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,768,0.016587555408477783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,1536,0.022257778379652236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,1536,0.024308444725142583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,512,0.010048000348938836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,256,0.006614222294754452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,512,0.014511111709806653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,768,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,256,0.010430222584141625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,128,0.0057502223385704895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,256,0.012520888613329994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,512,0.008023111356629265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,64,0.005668444352017508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,128,0.012422222230169507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,32,0.005642666584915585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,512,0.012088889049159156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,64,0.009739555418491364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,32,0.010055111514197456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,1024,0.019191111127535503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,1536,0.014932443698247274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,65536,0.40568889511956113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,128,0.009700444837411245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,65536,0.5627057817247179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,65536,0.4768853187561035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,51200,0.45091289944118923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,16384,0.10074933369954427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,51200,0.3731537659962972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,16384,0.1427217721939087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,16384,0.1264586713578966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,128,0.009760888914267222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,12288,0.10533066590627034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,51200,0.15219110912746853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,10240,0.06451200114356147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,12288,0.09624177879757351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,256,0.009722666607962714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,16384,0.050753778881496854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,10240,0.08270755741331312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,8192,0.07139288716846041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,12288,0.04033689035309686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,8192,0.026540444956885442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,8192,0.06833333439297147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,7168,0.04533511069085863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,10240,0.08891199694739448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,8192,0.04948177933692932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,6144,0.03748088743951585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,7168,0.06288622485266791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,7168,0.060814221700032554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,6144,0.0536133348941803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,5120,0.030284444491068523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,10240,0.03530133432812161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,65536,0.19461777475145128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,7168,0.024088889360427856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,5120,0.04621422290802002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,5120,0.01885333326127794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,4096,0.02510044475396474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,5120,0.04634311132960849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,6144,0.05484444565243191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,4096,0.016551110479566786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,3584,0.022965333527988855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,6144,0.02137600051032172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,12288,0.08445511261622111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,3584,0.015164444843928019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,51200,0.2836879889170329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,3584,0.035511109564039446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,3072,0.03029511041111416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,3072,0.014200000299347771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,3072,0.03122133347723219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,4096,0.038359112209743924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,2560,0.02647022240691715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,2048,0.015005333556069268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,2560,0.01309866706530253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,3072,0.01998222205373976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,3584,0.034266667233573064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,1536,0.011940444509188334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,2048,0.012118221984969245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,2048,0.02267644471592373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,2048,0.023776888847351074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,1536,0.011057777537239922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,1536,0.018544889158672757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,1536,0.01959111127588484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,768,0.008341333104504479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,1024,0.01015911085738076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,1024,0.01442577772670322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,4096,0.03939111034075419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,512,0.006648889018429651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,768,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,768,0.012255999777052136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,768,0.014507555299335055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,256,0.005314666777849197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,512,0.012459555433856117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,1024,0.009376889301670922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,2560,0.017969777186711628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,128,0.00490399988161193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,256,0.009027555584907532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,256,0.00904622260067198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,256,0.011406222151385413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,128,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,128,0.008393777741326226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,2560,0.02757155564096239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,32,0.004557333472702238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,64,0.008361777497662438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,65536,0.28298666742112905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,65536,0.4434666633605957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,65536,0.4663919872707791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,65536,0.1784515513314141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,512,0.009389333426952362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,512,0.010519111322032081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,51200,0.14262399408552381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,16384,0.0740977790620592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,64,0.004226666771703296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,128,0.011061333119869232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,16384,0.11439911524454753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,51200,0.21964089075724283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,32,0.00869777798652649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,12288,0.08700088659922282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,51200,0.3469031121995714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,12288,0.03847288754251268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,1024,0.016525333126386006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,16384,0.04872177706824409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,10240,0.08094933297899035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,8192,0.03685599896642897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,12288,0.09485599729749893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,8192,0.05881422095828586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,10240,0.04552622305022346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,51200,0.36683021651373965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,7168,0.031726221243540444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,8192,0.025394666526052687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,8192,0.06687288814120822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,10240,0.0725404421488444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,6144,0.027223111854659185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,16384,0.12377066082424587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,7168,0.060044447580973305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,10240,0.033599111768934466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,5120,0.023560000790490046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,6144,0.020566221740510728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,6144,0.045351111226611666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,4096,0.019516444868511625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,5120,0.01811733345190684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,5120,0.038605332374572754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,5120,0.04579377836651272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,4096,0.01585422290696038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,4096,0.0322088897228241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,4096,0.0377493335141076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,7168,0.023111111587948267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,12288,0.05557599994871351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,3072,0.015695111619101632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,3584,0.02883466747072008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,3584,0.03380977776315477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,2560,0.013723555538389417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,3072,0.026143999563323125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,6144,0.052137778864966504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,2560,0.02238133384121789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,3584,0.01828622155719333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,2048,0.011418666276666852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,2560,0.012832889126406776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,7168,0.05183466606669956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,3584,0.014936000108718872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,1536,0.009355555805895064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,2048,0.023215111758973863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,3072,0.030300445026821558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,1536,0.016002666619088914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,1024,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,1536,0.019315555691719055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,2560,0.026370666093296472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,1024,0.012447111308574677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,1024,0.010038221875826517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,1024,0.016183111402723525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,2048,0.011810666984981961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,3072,0.013832000394662222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,512,0.00591466658645206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,768,0.014091556270917257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,1536,0.010943999720944298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,256,0.005026666654480828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,512,0.012472888661755456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,768,0.0069511110583941145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,2048,0.01937333411640591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,128,0.004416888786686791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,256,0.01074044406414032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,768,0.010748444332016839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,128,0.008685333033402761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,128,0.007367110914654202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,64,0.004226666771703296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,128,0.010797333386209277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,64,0.007297777467303806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,256,0.008010666403505538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,65536,0.2148240009943644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,256,0.008722667064931657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,768,0.009737778041097853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,51200,0.17491377724541557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,512,0.00907733374171787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,65536,0.3235689004262288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,32,0.007353777686754863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,51200,0.14147999551561144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,51200,0.2539164490169949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,16384,0.09553777509265476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,65536,0.17652889092763266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,16384,0.08768977721532185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,16384,0.04878133204248217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,51200,0.28899555736117893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,16384,0.05355822377734714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,12288,0.0680933329794142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,12288,0.03821422325240241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,65536,0.37197065353393555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,12288,0.04127111037572225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,8192,0.027476444840431213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,12288,0.07238399982452393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,8192,0.04969244533114963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,10240,0.0358862214618259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,7168,0.02276266614596049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,8192,0.02504088812404209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,8192,0.04792355497678121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,10240,0.06089511182573107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,7168,0.04388622111744351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,7168,0.02273155583275689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,7168,0.04302666584650675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,6144,0.02048000031047397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,6144,0.038556443320380315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,5120,0.017310221989949543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,6144,0.038091556893454656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,5120,0.018340445227093168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,5120,0.0329022208849589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,4096,0.014360000689824423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,5120,0.03295466634962294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,6144,0.02014222244421641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,10240,0.05784355269538032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,3584,0.013177778157922955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,4096,0.02794044547610813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,3584,0.014826666977670459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,3072,0.011615110768212212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,3584,0.025790222816997107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,3072,0.02185955478085412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,3072,0.022856889499558344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,3072,0.013849777479966482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,4096,0.027496889233589172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,2560,0.019159111711714003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,3584,0.02443466583887736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,2048,0.00868444475862715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,2560,0.012892444100644855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,2048,0.01658933361371358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,2048,0.01789866718981001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,512,0.009316444396972656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,2048,0.011796444654464722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,10240,0.03375466664632162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,1536,0.01072266697883606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,1536,0.013841778039932251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,1536,0.01532444523440467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,2560,0.020298666424221463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,1024,0.010016000105275048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,768,0.005631111148330901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,32,0.00424177779091729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,1024,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,1024,0.010718222293588849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,768,0.011396444506115384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,512,0.0046728890803125174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,768,0.009658666948477427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,1024,0.012924444344308643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,4096,0.01614755557643043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,256,0.0042275554604000514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,512,0.00907911111911138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,512,0.008364444805516137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,512,0.01072533345884747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,128,0.003818666769398583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,256,0.008707555631796518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,256,0.006981333096822103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,256,0.009736888938479954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,64,0.003565333369705412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,128,0.0069057775868309875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,128,0.00906577789121204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,64,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,2560,0.010382222632567087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,32,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,65536,0.1676808860566881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,768,0.009390222529570261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,51200,0.12454222308264838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,65536,0.3194177680545383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,128,0.008660444782839881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,65536,0.31569867663913304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,32,0.003531555748648114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,51200,0.10623111327489217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,51200,0.2481671174367269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,1536,0.00758755538198683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,12288,0.032510221004486084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,16384,0.08605955706702338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,12288,0.06438310941060384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,12288,0.06679466697904798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,16384,0.043712000052134194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,51200,0.25082045131259495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,10240,0.05350311266051399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,8192,0.02048000031047397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,10240,0.025641777449183997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,16384,0.08549244536293878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,10240,0.05686666568120321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,7168,0.018216888109842937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,8192,0.018775110443433125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,8192,0.04354133208592733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,12288,0.029112001260121662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,8192,0.047424889273113675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,6144,0.015815999772813585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,7168,0.01705511079894172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,7168,0.042062222957611084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,6144,0.03415200114250183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,5120,0.014008000493049622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,6144,0.03746044304635789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,16384,0.03754577702946133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,5120,0.014198222094111972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,5120,0.02917511264483134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,5120,0.0319164428446028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,65536,0.13216533925798205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,4096,0.024156444602542456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,3584,0.010739555789364709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,4096,0.02700800034734938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,6144,0.015783111254374187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,3584,0.021912000245518152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,3584,0.02510844502184126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,4096,0.011795555551846823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,3072,0.022359111242824133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,2560,0.008773333496517604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,4096,0.012510221865442065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,10240,0.024736000431908503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,2560,0.01072799993885888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,3072,0.009736888938479954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,3072,0.019768888751665752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,2048,0.015049777097172208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,3072,0.011352889239788055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,1536,0.006907555378145642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,7168,0.038655112187067665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,2560,0.017328888177871704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,2048,0.007616889145639207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,1024,0.005571555760171678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,1024,0.010041777458455827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,2048,0.010316444767846001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,2048,0.017550221747822232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,1024,0.009034666750166152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,768,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,1024,0.01275644451379776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,2560,0.019688000281651814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,512,0.004516444272465176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,768,0.011234666738245221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,512,0.007713778151406183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,512,0.010390222072601318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,256,0.00423911131090588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,3584,0.011773332953453064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,1536,0.012449777788586087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,256,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,1536,0.009723555710580613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,128,0.004056888735956616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,256,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,256,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,64,0.003539555602603488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,128,0.007879111501905654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,128,0.009000000026490953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,512,0.008031999899281396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,64,0.006119110931952794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,65536,0.10335110955768162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,768,0.00887111077706019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,128,0.006049777898523543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,32,0.0038684445122877755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,65536,0.13078667057885066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,51200,0.08157244655821058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,32,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,1536,0.014818666709793938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,51200,0.21666933430565727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,51200,0.2382355531056722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,65536,0.27471910582648384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,16384,0.07228800323274401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,65536,0.3023742304907905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,16384,0.036864889992607966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,16384,0.08209511306550768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,10240,0.021328000558747187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,12288,0.028352889749738906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,12288,0.0549777779314253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,768,0.00870666652917862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,8192,0.019296889503796894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,10240,0.025386666258176167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,10240,0.054119110107421875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,8192,0.03821422325240241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,7168,0.01627288924323188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,12288,0.02678666677739885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,7168,0.03421777817938063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,7168,0.039820445908440485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,6144,0.015173332558737861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,16384,0.0313991109530131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,6144,0.030032889710532293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,8192,0.018603555030292936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,6144,0.034965333011415266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,5120,0.012770666844315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,6144,0.015563555889659457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,5120,0.014164444473054675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,5120,0.025435555312368605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,4096,0.009043555292818282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,5120,0.029909332593282063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,7168,0.017106667160987854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,3584,0.008372444245550368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,4096,0.021328889661365088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,4096,0.026016000244352553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,8192,0.04490310947100321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,3072,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,12288,0.06341511011123657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,51200,0.10515022277832031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,4096,0.012448000411192576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,2560,0.007264888948864407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,3072,0.011404444773991903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,3072,0.021314667330847845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,3584,0.019532443748580087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,2560,0.010757333702511258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,3584,0.023608888188997906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,2560,0.01867466668287913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,2048,0.010211555494202508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,2048,0.01349955548842748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,1536,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,10240,0.04645599921544393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,1536,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,1536,0.011505777637163797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,1024,0.004593777573770947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,1536,0.014367999302016364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,3584,0.011800888511869641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,2048,0.006102222121424145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,1024,0.009025777379671732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,1024,0.012077333198653327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,768,0.00869066682126787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,512,0.0038782221575578055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,768,0.011081777513027191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,512,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,512,0.010054222411579555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,768,0.004260444392760594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,256,0.0033039999090962936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,256,0.006261333409282897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,768,0.008054222497675154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,128,0.0032008888406885993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,256,0.007677333222495184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,128,0.00565244464410676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,128,0.008832889298597971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,2560,0.015449777245521545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,64,0.005623999983072281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,2048,0.01651733285850949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,32,0.003208000005947219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,65536,0.07128889030880399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,32,0.005640000104904175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,1024,0.009005332986513773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,65536,0.23688801129659018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,51200,0.059531556235419385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,65536,0.2998773256937663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,65536,0.13011199898189968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,256,0.00924622184700436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,16384,0.023714666565259297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,51200,0.1857057809829712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,51200,0.23589955435858834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,16384,0.062215997113121875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,64,0.0035182221068276297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,12288,0.01418133411142561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,16384,0.03671733207172818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,3072,0.01758311192194621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,12288,0.047968000173568726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,10240,0.012791111237472959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,12288,0.0625422199567159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,51200,0.10456177923414443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,10240,0.025602666868103877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,10240,0.04013777772585551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,8192,0.013097777962684631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,16384,0.08119110928641425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,8192,0.01863555610179901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,8192,0.032904889848497175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,8192,0.04438222116894192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,12288,0.029824889368481104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,7168,0.02917333443959554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,7168,0.018014222383499146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,512,0.008051555189821456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,6144,0.02587555514441596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,7168,0.011741333537631564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,6144,0.016561778055297006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,6144,0.034282667769326106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,5120,0.014840000205569796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,5120,0.022329777479171753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,4096,0.009280000295903947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,5120,0.029550221231248643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,7168,0.0392159985171424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,3584,0.008447111480765873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,128,0.007808888951937358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,4096,0.018927999668651156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,4096,0.025286222497622173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,3072,0.007718222008811102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,3584,0.023286221755875483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,3072,0.015776889191733468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,3072,0.02103022237618764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,6144,0.011591999895042844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,2560,0.0063440000845326324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,4096,0.01312711089849472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,2560,0.011063999599880643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,2560,0.013878222141000958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,3584,0.011865778101815118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,5120,0.010363555616802638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,2048,0.012152888708644442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,2048,0.016544888416926067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,3072,0.0116462219092581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,10240,0.05348711212476095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,1536,0.009685333404276106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,2048,0.0058444444504049085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,2048,0.010306666294733683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,1536,0.004940444396601783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,1024,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,768,0.003896000070704354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,1024,0.012140444583363004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,1536,0.014152000347773233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,3584,0.01720888912677765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,512,0.0038782221575578055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,768,0.00868711123863856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,1024,0.008395555946562026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,512,0.006666666517655055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,512,0.009712888962692684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,1536,0.010402667025725046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,256,0.00564533347884814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,2560,0.018395556343926322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,768,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,128,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,128,0.005292444593376583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,128,0.008389333056079017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,512,0.008115555677149031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,64,0.003169777699642711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,128,0.007637333538797166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,32,0.003181333343187968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,65536,0.06788444519042969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,64,0.005245333330498801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,32,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,256,0.007672000262472365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,768,0.010755555497275459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,16384,0.019466666711701285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,12288,0.01629599928855896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,51200,0.18520711527930367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,16384,0.0623191131485833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,10240,0.014589332871966891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,12288,0.04762755499945747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,10240,0.039897776312298246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,7168,0.012120889292822944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,51200,0.05543022354443868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,6144,0.01111288865407308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,65536,0.23664000299241808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,7168,0.029135998752382066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,6144,0.02567911148071289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,4096,0.007851555943489075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,8192,0.01273155543539259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,5120,0.022359111242824133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,4096,0.01885066595342424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,8192,0.03264710969395108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,1024,0.004565333326657613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,2560,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,256,0.008681777450773451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,2048,0.0053102220926019884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,3072,0.015505777464972602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,1536,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,2048,0.01203555530971951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,3584,0.007344000041484833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,1024,0.004239999999602635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,1536,0.010407110883129967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,768,0.003891555385457145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,3584,0.017252445220947266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,512,0.003543111185232798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,1024,0.00850133349498113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,256,0.0032035555276605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,768,0.007621333003044128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,128,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,2560,0.013782222237851886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,64,0.0028568889117903183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,256,0.003540444291300244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,32,0.0028479999552170434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,128,0.0052764444715446895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,64,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,65536,0.06713689035839505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,51200,0.05132266547944811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,32,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,16384,0.016893333858913846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,65536,0.24323201179504395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,5120,0.00904177791542477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,12288,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,51200,0.18506311045752633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,10240,0.009694221946928237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,8192,0.008997333546479544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,256,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,7168,0.008344000412358178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,12288,0.04752177662319607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,6144,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,8192,0.032446222172843084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,5120,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,7168,0.02902488907178243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,6144,0.025777777036031086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,3584,0.008361777497662438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,5120,0.022264000442292955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,3072,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,16384,0.06340355343288846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,3584,0.017267554998397827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,3072,0.015544000599119397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,2048,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,2560,0.01386755539311303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,1536,0.004612444589535396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,4096,0.007291555404663086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,1024,0.0038862224254343244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,2048,0.012080888781282636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,768,0.0038746665749284956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,1536,0.010455999937322406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,512,0.0035573335157500375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,4096,0.018925334016482036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,1024,0.008387555678685507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,2560,0.006754666566848755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,768,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,512,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,10240,0.03991733325852288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,256,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,32,0.0028737777223189673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,3072,0.006595555692911148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,128,0.005446222093370225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,64,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,32,0.004973333328962326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,512,0.00665511108107037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,128,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,256,0.0032986667421129015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,64,0.0028657778683635923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,16384,1.2383964326646593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,12288,2.68120977613661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,12288,0.8956178029378256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,16384,3.5755848354763455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,12288,2.8312791188557944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,16384,4.0698161655002165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,10240,0.7599715656704373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,51200,4.526888953314887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,8192,1.7710480160183375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,10240,2.184612486097548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,8192,0.5995359950595432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,10240,2.507911046346029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,8192,1.8077138264973958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,7168,0.5157697995503744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,6144,0.4759990904066298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,6144,1.444655948215061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,7168,1.5592515733506944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,5120,1.2441138161553276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,5120,0.3892515500386556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,4096,0.8571182356940376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,5120,1.1359422471788194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,7168,1.7603510750664606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,4096,0.8865049150254992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,3584,0.7850249078538684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,3584,0.7642986509535047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,3072,0.6593697865804037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,4096,0.32058578067355686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,3072,0.264810668097602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,3072,0.6834595468309191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,6144,1.3696186277601454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,2560,0.5960293345981175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,3584,0.2991733286115858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,2560,0.23147734006245932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,2560,0.5622222158643935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,2048,0.18937244680192736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,1536,0.35628000895182294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,2048,0.4599279827541775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,1536,0.36341333389282227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,51200,11.093847486707899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,1024,0.1221804486380683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,1024,0.27260533968607586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,768,0.2216693295372857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,2048,0.41140267584058976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,768,0.10807289017571343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,512,0.15720533000098333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,768,0.21667022175259062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,256,0.1018257803387112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,1024,0.2485040028889974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,128,0.0792906681696574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,256,0.1675715578926934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,256,0.08577422300974528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,128,0.08283022377226087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,64,0.06837333573235406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,128,0.15771467155880398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,512,0.09576621982786392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,1536,0.15098933378855386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,32,0.16274221738179526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,32,0.07158222463395861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,64,0.16042310661739773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,512,0.186246222919888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,51200,16.656534830729168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,16384,2.763348473442925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,16384,0.8884275754292806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,51200,3.5055421193440757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,65536,4.574516296386719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,16384,3.234465705023872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,12288,0.696108447180854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,12288,2.292453342013889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,12288,2.3703216976589627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,10240,1.8952354855007594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,51200,8.921858893500433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,10240,0.573911984761556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,10240,1.9008079104953344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,8192,1.5027341842651367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,8192,1.4246062172783747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,7168,0.42029423183865017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,7168,1.526691542731391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,7168,1.215571509467231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,8192,0.48218843672010636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,6144,0.4174826674991184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,6144,1.0888755586412218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,5120,0.8501608636644152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,6144,1.068161752488878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,5120,0.30773067474365234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,65536,11.293301052517362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,5120,0.8744604322645398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,4096,0.26112622684902614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,3584,0.5990364286634657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,4096,0.6974506907992892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,3072,0.5003386603461372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,3072,0.20252354939778647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,3584,0.5995564460754395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,2560,0.4400471051534017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,2560,0.1850293344921536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,3584,0.2346515655517578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,2048,0.35081688563028973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,2560,0.4387386639912923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,2048,0.15416533417171902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,2048,0.36195821232265896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,1536,0.12309866481357151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,4096,0.7028409110175239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,1024,0.19782400131225586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,768,0.15200977855258518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,51200,13.430518256293404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,1024,0.2150071064631144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,512,0.1096080011791653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,1536,0.24899466832478842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,512,0.07663644684685601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,512,0.147180438041687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,3072,0.5345190895928277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,256,0.07762489053938124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,128,0.062246223290761314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,256,0.06854222218195598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,1024,0.09539110793007745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,256,0.13247288597954646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,64,0.05944177839491102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,128,0.06684888733757867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,32,0.06324622366163465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,768,0.17110933197869194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,64,0.12688622209760878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,32,0.12881510787540013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,1536,0.28604621357387966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,65536,18.39142015245226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,768,0.08407911327150132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,65536,1.4614551332261827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,16384,0.8617351320054797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,51200,1.1079848607381184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,128,0.12455288569132487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,16384,0.30030043919881183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,12288,0.23082311948140463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,12288,0.9138541751437717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,65536,3.696072048611111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,16384,1.0518169403076172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,10240,0.5562622282240126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,10240,0.1970097753736708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,8192,0.4937538041008844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,10240,0.62700531217787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,51200,2.899271011352539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,8192,0.1752533382839627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,51200,4.10993406507704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,12288,0.7105049027336968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,7168,0.42819645669725204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,7168,0.1440933280520969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,6144,0.35712265968322754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,6144,0.1598844395743476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,5120,0.3179679976569282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,6144,0.3448622226715088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,5120,0.13845154974195692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,4096,0.2352177831861708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,5120,0.33044976658291286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,4096,0.09557244512769912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,8192,0.47260088390774196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,4096,0.2286773257785373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,3584,0.0788559979862637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,3072,0.16876799530453154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,3584,0.20368977387746176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,2560,0.1489528947406345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,3072,0.17537332905663383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,2048,0.11530044343736436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,65536,5.669619666205512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,1536,0.0874933335516188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,2560,0.14995466338263616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,2048,0.12416533629099528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,1536,0.043805334303114146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,1536,0.09870844417148167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,7168,0.4031377898322211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,3072,0.07031466563542683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,768,0.052503999736573964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,1024,0.0761022236612108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,2560,0.062168889575534396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,512,0.03944977786805894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,768,0.06048088603549533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,256,0.02643999954064687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,3584,0.2019644445843167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,256,0.028260442945692275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,512,0.0521395570702023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,128,0.025311999850802954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,256,0.046642667717403836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,128,0.027691556347741023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,1024,0.03666399916013082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,64,0.02048000031047397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,128,0.04407111141416761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,32,0.020985777179400127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,768,0.033858666817347206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,64,0.04493066668510437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,32,0.045759111642837524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,2048,0.05436088972621494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,1024,0.06852622164620294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,16384,0.6912195417616102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,65536,1.039231088426378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,51200,0.8452142079671224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,512,0.031331555710898504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,16384,0.2610630989074707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,51200,2.0426578521728516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,12288,0.17874577310350206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,12288,0.5317306518554688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,65536,2.727575937906901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,16384,0.8773848745557996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,10240,0.4124035570356581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,10240,0.15122932857937282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,8192,0.35928087764316136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,8192,0.12435021665361191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,7168,0.31899023056030273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,8192,0.3552089002397325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,12288,0.611385769314236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,6144,0.26739290025499135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,51200,3.540586683485243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,7168,0.39806400405036074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,65536,3.9939388699001737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,5120,0.2249493334028456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,5120,0.08222133583492704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,4096,0.17122933599683973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,5120,0.24624265564812553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,3584,0.15263377295600042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,4096,0.203167994817098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,3584,0.06320089101791382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,7168,0.11175199349721272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,3584,0.15955021646287706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,3072,0.0555902189678616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,6144,0.2636195553673638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,2560,0.0492195553249783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,2560,0.11611821916368271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,4096,0.06877155436409844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,2048,0.042465776205062866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,2048,0.09631199969185723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,6144,0.0989875528547499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,3072,0.13861421744028726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,1536,0.07653066847059461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,1024,0.048692444960276283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,1024,0.028905779123306274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,2048,0.09898489051394993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,768,0.03816711240344577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,3072,0.12645422087775335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,512,0.02938488788074917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,768,0.046365333928002246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,2560,0.10573689142862956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,1536,0.03470755616823832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,256,0.02144711050722334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,10240,0.4495840072631836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,1024,0.05793599949942695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,256,0.021918222308158875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,128,0.018569777409235638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,1536,0.06709333260854085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,64,0.015809777710172866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,128,0.021496888664033677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,128,0.03307999836073981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,512,0.024439111351966858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,32,0.03295822276009454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,256,0.03507733345031738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,768,0.026890666948424444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,32,0.015963556038008798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,512,0.03921688927544488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,64,0.03314933180809021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,65536,0.9186906814575195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,16384,0.6625075340270996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,51200,0.702089786529541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,16384,0.25108978483412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,65536,2.4857351515028214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,16384,0.6177528699239095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,12288,0.5163288646274143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,12288,0.16707289218902588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,10240,0.41711288028293186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,10240,0.14657689465416804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,12288,0.5290186670091417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,8192,0.3384666707780626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,10240,0.37183199988471133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,8192,0.10646400186750625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,7168,0.2731502321031359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,65536,3.2634826236300998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,7168,0.09285333421495225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,6144,0.23898400200737846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,8192,0.4045262336730957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,6144,0.0831893351342943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,5120,0.19405599435170492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,6144,0.24259466595119902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,5120,0.07008177704281278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,5120,0.1963662174012926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,3584,0.14081066184573704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,4096,0.14911111195882162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,3584,0.052389333645502724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,3584,0.13223911656273737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,51200,2.6124293009440103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,51200,1.8364285363091364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,2560,0.09041600094901191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,4096,0.15082666609022352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,2560,0.04156000084347195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,4096,0.05961600277158949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,2560,0.09875288936826919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,2048,0.036222222778532244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,2048,0.08225778076383802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,1536,0.02974489000108507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,1536,0.06527022520701091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,1024,0.0403733319706387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,1024,0.024499555428822834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,2048,0.07437866926193237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,768,0.035402665535608925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,768,0.022263111339675054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,1536,0.059916443294949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,512,0.02643199927277035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,512,0.020296888218985665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,3072,0.04667822188801236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,7168,0.27828089396158856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,256,0.018836443622907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,128,0.016374222106403775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,1024,0.04956000049908956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,128,0.018591110905011494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,3072,0.10898666911655003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,64,0.015062222878138224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,32,0.01943288909064399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,256,0.01937866707642873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,64,0.027823110421498615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,32,0.027469333675172593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,256,0.029330667522218492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,128,0.027789334456125896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,768,0.03997866643799676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,65536,0.7615084648132324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,65536,1.8473705715603297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,51200,0.5779004626803929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,51200,1.5448267194959853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,16384,0.17839111222161186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,512,0.032736887534459434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,12288,0.40070756276448566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,12288,0.12564088238610163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,10240,0.3392968972524007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,12288,0.36667821142408585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,16384,0.5858826637268066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,10240,0.10731910996966892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,8192,0.25003822644551593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,10240,0.3059804439544678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,8192,0.08666666348775227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,3072,0.1161831087536282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,16384,0.5189333491855198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,7168,0.07730488644705878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,7168,0.21408889028761122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,65536,2.7253990173339844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,6144,0.06816266642676459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,51200,2.2839537726508246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,5120,0.05997777647442288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,5120,0.15418399704827204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,7168,0.1957857741249932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,4096,0.048559112681282886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,4096,0.12267288896772598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,3584,0.10457066694895427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,6144,0.18488710456424287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,3584,0.04355466696951124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,3072,0.08884177605311076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,3584,0.1088542209731208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,2560,0.07342044512430827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,3072,0.038797332180870905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,3072,0.09529511133829753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,2048,0.06069688664542305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,2560,0.03463733196258545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,1536,0.04847022228770786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,6144,0.19876621829138863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,2048,0.06845244434144762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,1024,0.03382577829890781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,5120,0.1657359997431437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,1536,0.054209778706232704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,768,0.02568000058333079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,1024,0.021011556188265484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,1024,0.04135911001099481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,512,0.02113155523935954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,768,0.018931556079122756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,8192,0.2407297823164198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,256,0.0155102229780621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,512,0.017151110702090792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,512,0.027628444963031348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,128,0.014106666048367819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,256,0.015974221958054435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,256,0.024088000257809956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,64,0.011879111329714457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,128,0.016136889656384785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,128,0.023403555154800415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,4096,0.13041510846879747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,768,0.03349422083960639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,2560,0.08147822486029731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,2048,0.029891557163662378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,1536,0.024923556380801733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,32,0.012074666718641916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,64,0.023277334041065637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,32,0.02304711110062069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,65536,1.6326915952894423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,51200,1.2292968961927626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,51200,0.4992026752895779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,16384,0.15065156088935003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,16384,0.43868266211615664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,12288,0.11632443798912896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,12288,0.3381182352701823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,65536,0.6412755648295084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,65536,2.4284710354275174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,51200,1.9458808898925781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,8192,0.22356621424357095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,10240,0.26866398917304146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,8192,0.08471733331680298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,8192,0.22829333941141763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,12288,0.31304089228312176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,6144,0.15392977661556667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,7168,0.18819555971357557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,6144,0.06146933635075887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,5120,0.14681066407097712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,6144,0.16422311464945474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,10240,0.09522310892740886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,4096,0.10353955295350815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,5120,0.13772711488935682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,16384,0.4336862299177382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,7168,0.19431289037068686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,4096,0.04309689005215963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,3072,0.07524444659550984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,4096,0.10972978009117974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,7168,0.07270489136377971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,3584,0.09700622161229451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,3072,0.036166220903396606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,3072,0.08528977632522583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,10240,0.23392976654900444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,2560,0.0321777794096205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,2560,0.07295822434955172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,5120,0.05338311195373535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,2048,0.02789333462715149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,2048,0.061379558510250516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,3584,0.08346044354968601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,1536,0.023215999205907185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,1536,0.048991110589769155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,768,0.024119110571013555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,1024,0.01947733263174693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,2560,0.06854044728808932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,1024,0.037231110864215426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,768,0.018027555611398485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,768,0.030327998929553564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,2048,0.050921777884165444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,256,0.014120888378885059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,512,0.016395555602179635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,512,0.024883555041419134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,128,0.011745777395036487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,256,0.015085332923465304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,256,0.021718222233984206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,1024,0.02846222122510274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,64,0.01037333326207267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,3584,0.03878133495648702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,32,0.010844444235165914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,512,0.017362667454613578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,64,0.020985777179400127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,32,0.0206977774699529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,1536,0.039489779207441546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,128,0.015150222513410779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,65536,0.5971431202358669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,65536,1.2758329179551866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,51200,1.1714098188612196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,16384,0.3764959971110026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,51200,0.47085777918497723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,16384,0.16095822387271458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,16384,0.401239103741116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,12288,0.12554932965172663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,128,0.020964443683624268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,12288,0.32741067144605845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,10240,0.0880204439163208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,10240,0.23602400885687935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,8192,0.07070400317509969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,8192,0.1869431071811252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,7168,0.16711910565694174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,12288,0.2800559997558594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,6144,0.13399377134111193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,7168,0.16586221588982478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,10240,0.1996257834964328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,6144,0.05603466431299845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,5120,0.1110959980222914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,51200,1.678615993923611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,4096,0.08937155538135105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,8192,0.18527021672990587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,5120,0.12069955137040879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,4096,0.039651556147469416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,4096,0.09603022204505073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,3072,0.06389066908094618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,3584,0.03513777918285794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,3584,0.08533244662814671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,7168,0.06517244709862603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,3072,0.03167288833194309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,3072,0.07515733109580146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,2560,0.028704888290829126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,65536,2.289665857950846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,2560,0.06512533293830024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,5120,0.046522667010625206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,1536,0.03484266665246751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,2048,0.024655110306209985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,3584,0.0812915563583374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,1024,0.025656888882319134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,1536,0.020929776959949072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,768,0.019279999865425956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,1024,0.017998221847746108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,1024,0.03294310967127482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,512,0.015309333801269531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,768,0.015860444969601102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,768,0.02725155486000909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,256,0.011756444142924415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,512,0.014751111467679342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,6144,0.14236533641815186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,512,0.02232711182700263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,256,0.013757333159446716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,256,0.019276445110638935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,2048,0.053845332728491895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,64,0.009011555876996782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,128,0.018791110979186166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,32,0.009350222018029954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,64,0.018568888306617737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,32,0.01851466629240248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,2048,0.04594755503866407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,128,0.010421333213647207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,65536,0.543514675564236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,1536,0.04324088825119866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,65536,1.385216924879286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,2560,0.05695022477043999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,51200,1.058183987935384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,51200,0.4408302307128906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,16384,0.11958400408426921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,65536,1.6321858300103083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,16384,0.4115004539489746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,51200,1.1917911105685763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,12288,0.0894213318824768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,10240,0.19006310568915474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,128,0.013850666582584381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,12288,0.23783466551038954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,8192,0.1606631146536933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,7168,0.13764088683658177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,8192,0.16120089424981013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,16384,0.3141911029815674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,7168,0.055445333321889244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,7168,0.1445360051261054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,5120,0.097488886780209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,12288,0.256510231229994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,6144,0.12252533435821533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,5120,0.04113599989149306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,5120,0.10401155551274617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,10240,0.08245244291093615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,3584,0.06820088624954224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,8192,0.06091377470228407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,4096,0.08243822389178805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,6144,0.11559377776251899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,3584,0.030557334423065186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,3072,0.06257955233256023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,3072,0.06475644641452365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,3072,0.028036445379257202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,2560,0.04850844542185465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,10240,0.2005315621693929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,2560,0.025075554847717285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,2048,0.04031377699640062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,2560,0.05596888727611966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,1536,0.030128889613681372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,2048,0.021903999977641638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,2048,0.046776887443330556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,1024,0.02142133315404256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,1536,0.018277333842383493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,1536,0.037470221519470215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,768,0.017479999197853934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,4096,0.07861244678497314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,1024,0.015499555402331881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,512,0.014188443620999655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,768,0.014582221706708273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,4096,0.03433777888615926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,768,0.023631999890009563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,512,0.01352355546421475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,512,0.019620445039537217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,128,0.009327111144860586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,256,0.017160000072585214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,64,0.008906666603353288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,128,0.012727999852763282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,32,0.00868444475862715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,64,0.016213332613309223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,32,0.016228444046444364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,256,0.010599999792046018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,3584,0.07385066482755873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,6144,0.048215111096700035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,65536,0.8674160109625922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,128,0.016472889317406546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,1024,0.028559999333487615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,16384,0.24889511532253691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,51200,0.34279823303222656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,16384,0.09731022516886394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,65536,1.3375946680704753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,12288,0.19224177466498482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,256,0.012708444562223224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,16384,0.26890399720933705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,10240,0.17688443925645617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,65536,0.4139520062340631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,10240,0.06531555785073175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,10240,0.1670151154200236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,8192,0.05207644568549263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,51200,0.6228284305996364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,8192,0.13535111480289036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,7168,0.04641333222389221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,6144,0.09108089076148139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,7168,0.11762577957577175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,6144,0.04024089044994778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,12288,0.07841866546207003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,6144,0.10181866751776801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,12288,0.20054488711886934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,5120,0.035616000493367515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,5120,0.08673777845170762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,3584,0.05417244301901924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,7168,0.11840532885657416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,4096,0.0699146654870775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,3584,0.026965333355797663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,3584,0.06234133243560791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,5120,0.08300711048973931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,3072,0.024751999311976965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,3072,0.054770665036307446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,4096,0.06316177712546454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,2560,0.02219377789232466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,2560,0.047188444270028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,4096,0.03011466728316413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,1536,0.027705777022573683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,2048,0.03983911209636264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,3072,0.047000888321134776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,1024,0.017664000391960144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,1536,0.01707555519209968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,51200,1.054433822631836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,1024,0.01458666721979777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,1024,0.02421244482199351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,8192,0.14509244759877524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,2048,0.03499644332461887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,768,0.013461333182122974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,768,0.020453333854675293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,256,0.009112888740168678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,512,0.012432888978057437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,2048,0.019543111324310303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,512,0.016933333542611863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,128,0.007667555577225155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,256,0.011771555576059552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,256,0.014800889624489678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,128,0.011738667057620155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,32,0.007119999991522894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,128,0.014204444156752693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,768,0.013996443814701505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,64,0.014197332991494073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,512,0.011749332977665795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,65536,0.7679342163933648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,64,0.006796444455782573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,1536,0.03224622209866842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,2560,0.042209777567121715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,65536,1.1864257388644748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,16384,0.21852710511949328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,51200,0.29877599080403644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,16384,0.09352888663609822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,16384,0.23773156272040474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,65536,0.39087645212809247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,51200,0.9403022130330404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,10240,0.13951199584537083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,12288,0.06992444727155897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,12288,0.17750844690534803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,10240,0.05881777736875746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,10240,0.15899822447035047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,8192,0.04734755555788676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,32,0.014084445105658637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,7168,0.0984862248102824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,8192,0.12028977606031631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,6144,0.08855466710196601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,7168,0.04086933202213711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,12288,0.17007111178504097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,5120,0.06438577837414212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,7168,0.10702666971418594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,4096,0.057132442792256675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,5120,0.03158933255407546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,5120,0.07848355505201551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,8192,0.11976533465915257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,4096,0.02681511143843333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,3072,0.03907733493381076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,3584,0.02419822249147627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,3584,0.05727377865049574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,3072,0.022234666678640578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,2560,0.03474400109714932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,3072,0.049902220567067467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,2048,0.027660444378852844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,6144,0.03640888796912299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,2560,0.04320355587535434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,1536,0.022200889057583276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,2048,0.036264889770083956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,3584,0.04487466812133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,1024,0.017920000685585868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,4096,0.06311377551820543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,1536,0.02942755487230089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,768,0.014168888330459595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,1024,0.01352533366945055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,1024,0.022239999638663396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,512,0.011892444557613797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,768,0.0185208883550432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,6144,0.09223022063573201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,256,0.008919110728634728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,512,0.01568000018596649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,256,0.01144088887506061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,1536,0.015454222758611044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,256,0.013779555757840475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,128,0.01107377807299296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,64,0.006665777828958299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,768,0.012677333421177335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,32,0.006683555742104848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,2560,0.020250666472646926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,51200,0.6289928754170736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,64,0.01315822203954061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,32,0.013110222087966071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,128,0.00755555596616533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,65536,0.6828746795654297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,51200,0.5690079794989692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,51200,0.2711217668321398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,65536,1.0246569315592449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,512,0.011784888803958893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,51200,0.7566551102532281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,12288,0.14029688305324978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,16384,0.20878399742974174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,12288,0.05995821952819824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,65536,0.31469776895311147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,10240,0.11836889055040146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,12288,0.1568631066216363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,8192,0.0918240017361111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,10240,0.050832000043657094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,10240,0.13230489359961614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,16384,0.2022248903910319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,8192,0.10629600286483765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,2048,0.01760444376203749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,7168,0.0357582237985399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,6144,0.06916622320810954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,7168,0.09454222520192464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,6144,0.03163733416133457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,6144,0.08254400226804945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,5120,0.027809778849283855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,4096,0.04614311125543383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,5120,0.06894310977723864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,128,0.013090666797426013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,3584,0.040503111150529646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,4096,0.023796444137891132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,7168,0.08836088577906291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,4096,0.0561715563138326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,3584,0.021772444248199463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,3584,0.05053155620892843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,3072,0.020252444677882724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,16384,0.07732799980375502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,3072,0.044461333089404635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,5120,0.05659111340840658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,2560,0.017873777283562552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,2560,0.038789331912994385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,2048,0.01579466627703773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,1536,0.0192248887485928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,2048,0.03280355532964071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,1024,0.014504000544548035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,1536,0.01428711083200243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,1536,0.026597332623269822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,3072,0.03462488783730401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,1024,0.012413333687517377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,1024,0.019966221517986722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,2560,0.028452446063359577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,512,0.009379555781682333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,768,0.011748444702890186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,256,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,2048,0.023591111103693645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,512,0.01092177795039283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,128,0.006649777707126405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,256,0.010397333237859938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,256,0.012434666355450949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,64,0.005978666659858491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,128,0.010381333529949188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,128,0.011921777493423887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,32,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,8192,0.03997688823276096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,768,0.011502222054534487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,64,0.012005333271291522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,32,0.012106666962305704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,768,0.016875555117925007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,65536,0.5618266529507107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,65536,0.2884026633368598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,51200,0.4234044551849365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,16384,0.15851022137535944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,51200,0.22643023067050508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,16384,0.07147910859849718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,65536,0.7949164708455404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,16384,0.18358043829600015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,12288,0.05572888586256239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,8192,0.08066933022605048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,10240,0.04902311166127523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,10240,0.12231555249955918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,8192,0.03797244363360935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,7168,0.06646844413545397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,8192,0.09460266431172688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,6144,0.06079022089640299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,7168,0.03536355495452881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,7168,0.08387821912765503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,12288,0.11530666881137425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,5120,0.05025777882999844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,6144,0.03086044390996297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,6144,0.07298400004704793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,5120,0.026657778355810378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,12288,0.14044622580210367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,5120,0.06110577450858223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,3584,0.033596442805396184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,4096,0.022563555174403723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,4096,0.04983733428849114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,3584,0.04492177897029453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,3584,0.020455110404226515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,2560,0.025772444076008264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,3072,0.01885333326127794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,3072,0.03973866502443949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,51200,0.6208000183105469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,4096,0.039269334740108915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,2560,0.034734222624037005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,1536,0.017627555463049147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,2048,0.015523556205961438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,2048,0.02927644385231866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,1024,0.013295999831623502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,1536,0.013814222481515674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,1536,0.02346044447686937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,3072,0.02902311086654663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,768,0.011782222323947482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,1024,0.012137778103351593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,1024,0.018031999468803406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,768,0.011427555647161273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,768,0.015309333801269531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,10240,0.10189333226945664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,512,0.010668444136778513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,512,0.012828444441159567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,128,0.006030222194062338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,256,0.010095111197895473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,256,0.011339555184046427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,64,0.005614222337802251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,128,0.01073155552148819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,512,0.01424800025092231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,32,0.005923555543025334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,512,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,32,0.010729778144094678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,65536,0.395614226659139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,2048,0.02221155497762892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,51200,0.3079564571380615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,128,0.01013422260681788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,65536,0.6395084593031142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,16384,0.11613689528571235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,64,0.010755555497275459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,16384,0.06517244709862603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,12288,0.09735555781258477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,16384,0.15585600005255804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,2560,0.01722044414944119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,10240,0.08100622230105929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,12288,0.11939556068844265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,10240,0.042371554507149585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,10240,0.0988044473859999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,8192,0.06602133644951715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,51200,0.5063591003417969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,7168,0.055870221720801465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,8192,0.0340924428568946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,8192,0.08038222127490573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,7168,0.030914667579862807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,7168,0.07106488943099976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,12288,0.04883111185497708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,5120,0.04305511050754123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,6144,0.027098667290475633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,6144,0.06168711185455322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,4096,0.031192888816197712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,5120,0.023761777414215937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,5120,0.05173155665397644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,3584,0.028510222832361858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,4096,0.04253422220547994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,6144,0.048251556025611035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,3584,0.018696000178654987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,3072,0.023358222511079576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,3584,0.038374221987194486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,2560,0.021661332911915247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,3072,0.01705511079894172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,3072,0.03376977642377218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,2048,0.017235555582576327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,2560,0.029013332393434312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,51200,0.20862578021155465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,1536,0.013790222505728403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,2048,0.014431110686726041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,1024,0.010098666780524785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,1536,0.012476444244384766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,1536,0.020689777202076383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,768,0.008691555923885768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,1024,0.010802666346232096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,1024,0.01568622224860721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,512,0.007184888753626082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,768,0.01038933296998342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,768,0.013409777647919126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,2560,0.015467555986510383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,256,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,512,0.009720000127951304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,512,0.01143200033240848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,128,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,2048,0.02491644521554311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,64,0.004902222090297275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,128,0.009351111120647853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,32,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,64,0.009349333743254343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,32,0.009719111025333405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,65536,0.37522133191426593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,4096,0.020114666885799833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,256,0.010001777774757808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,65536,0.18977777163187662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,51200,0.23650754822625053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,256,0.007634667058785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,16384,0.09502044651243423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,51200,0.14818044503529867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,12288,0.06888444556130303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,51200,0.4161048995123969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,16384,0.1323644452624851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,10240,0.059260441197289355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,256,0.009394666386975182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,12288,0.10095644659466213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,8192,0.04699644446372986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,10240,0.03495111068089803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,10240,0.08361066712273492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,128,0.009420444567998251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,7168,0.03886933459175958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,8192,0.02589599953757392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,65536,0.25244800249735516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,7168,0.023338667220539514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,16384,0.049923555718527905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,7168,0.059749331739213735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,5120,0.029109332296583388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,12288,0.03811111052831014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,5120,0.018637332651350234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,4096,0.023365333676338196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,5120,0.04433155390951368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,3584,0.021504888931910198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,4096,0.016207999653286405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,65536,0.5261448754204644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,3584,0.015017777681350708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,3072,0.019056888090239633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,8192,0.06830933358934191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,2560,0.015997333659066092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,3072,0.013927999469969006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,6144,0.021312889125612047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,6144,0.05196977655092875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,2048,0.013880000346236758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,2048,0.01166044423977534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,1536,0.01144177797767851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,2048,0.021591999464564856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,4096,0.03698755635155572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,1024,0.008947555389669206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,1536,0.010760889285140567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,6144,0.03487288951873779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,1024,0.009744889206356471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,768,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,1024,0.013897778259383308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,2560,0.012834666503800286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,512,0.006653333289755716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,768,0.011911110745535957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,512,0.009023110899660323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,512,0.010241777532630498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,3584,0.03359288970629374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,256,0.009001777403884465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,256,0.008555555509196388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,2560,0.025235555238193933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,128,0.008368000388145447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,128,0.00870311094654931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,256,0.0053048887186580235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,32,0.004577777865860197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,64,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,1536,0.018030222919252183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,32,0.008325333396593729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,65536,0.2766791184743245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,768,0.009462222456932068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,51200,0.20284622245364717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,65536,0.1732355621125963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,65536,0.39639467663235134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,64,0.004571555389298333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,51200,0.13810400168100992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,16384,0.04689955711364746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,16384,0.10353510909610325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,12288,0.03668444355328878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,12288,0.07755377557542589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,128,0.004899555610285865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,8192,0.0333280000421736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,10240,0.06497155295477973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,16384,0.06409511301252577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,7168,0.028455999162462022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,12288,0.05083733465936449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,51200,0.309846215777927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,6144,0.02374222212367588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,7168,0.02254399988386366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,10240,0.044433779186672635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,5120,0.021634666456116572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,6144,0.019911110401153564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,3072,0.029167999823888142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,6144,0.0409840014245775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,4096,0.017216889394654166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,5120,0.017833777599864535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,3584,0.015341333217091031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,4096,0.015582222077581616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,8192,0.02461155586772495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,3072,0.01348444405529234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,3584,0.02609955602222019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,3072,0.013824000126785703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,2560,0.012281777958075205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,10240,0.03261333372857835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,2560,0.020419556233617995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,2560,0.012614222036467658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,4096,0.029153777493370905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,8192,0.052878220876057945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,1536,0.008700444466537898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,2048,0.011446221835083432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,2048,0.01742488808102078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,1024,0.007000889215204451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,1536,0.010605333579911126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,1536,0.014518222875065274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,1024,0.009743111001120673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,1024,0.01145244472556644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,5120,0.03482755687501695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,768,0.009351111120647853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,3584,0.014728888869285583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,512,0.008730666504965888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,256,0.004580444345871608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,7168,0.04666755596796671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,768,0.0064311110311084324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,128,0.0038853333228164246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,2048,0.010399111443095738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,128,0.00702311098575592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,64,0.003887999802827835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,512,0.008744888835483128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,32,0.0038524443904558816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,3072,0.02331111166212294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,64,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,256,0.008540444076061249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,65536,0.20970933967166475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,256,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,51200,0.16767111089494494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,65536,0.16893155044979521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,65536,0.3450302282969157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,128,0.008397333323955536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,51200,0.27526667382982045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,12288,0.04057244459788004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,16384,0.04651022288534376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,16384,0.09094488620758057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,32,0.007287999822033777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,10240,0.03394133183691237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,12288,0.03597955571280585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,512,0.005410666680998272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,10240,0.032413333654403687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,10240,0.05755822526084053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,51200,0.137809779908922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,8192,0.024882665938801233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,8192,0.047000888321134776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,7168,0.021879111727078755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,6144,0.021503110726674397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,12288,0.06855199734369914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,6144,0.019941333267423842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,6144,0.03667199942800734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,5120,0.01882933411333296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,16384,0.05273955398135715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,4096,0.01531022290388743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,5120,0.01775733298725552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,5120,0.031200889084074233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,3584,0.014304889572991265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,7168,0.0420720014307234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,3584,0.014490667316648694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,3072,0.012536889149083031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,768,0.010043555663691627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,3072,0.013783999615245394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,2560,0.010815111299355825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,4096,0.015569777952300178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,4096,0.026106667187478807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,2048,0.009342222577995723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,3584,0.023521777656343248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,1536,0.008603555460770925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,8192,0.02739022175470988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,1536,0.010506666368908353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,1024,0.006984000404675801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,1024,0.010737777584128909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,1024,0.009745777481132084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,768,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,3072,0.021200888686709937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,768,0.009345778160625035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,512,0.005244444641802046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,2560,0.01867288847764333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,512,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,512,0.008737777670224508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,2048,0.011436444189813403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,2048,0.015878222054905362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,128,0.003933333274390963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,256,0.007002666592597961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,7168,0.023727999793158636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,64,0.0035457776652442086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,128,0.008374222450786168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,768,0.009377777576446533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,32,0.003589333345492681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,65536,0.14658844470977783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,32,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,256,0.004249777644872665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,51200,0.12920533286200628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,65536,0.12743555174933538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,65536,0.30168355835808647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,1536,0.013393777940008374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,16384,0.040992889139387347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,51200,0.10184977451960246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,12288,0.03038933210902744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,51200,0.23783466551038954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,16384,0.08089155620998807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,2560,0.012538666526476542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,12288,0.02770400047302246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,12288,0.060009777545928955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,8192,0.01941688855489095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,10240,0.02475999957985348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,10240,0.05063822203212314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,256,0.008456889126035903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,7168,0.017872000734011333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,8192,0.041343109475241765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,6144,0.01573244399494595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,7168,0.01676177812947167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,128,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,7168,0.036981334288915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,6144,0.015426667200194465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,6144,0.03257333238919576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,5120,0.013793778088357715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,5120,0.027776888675159876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,4096,0.012443555725945367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,8192,0.0181022220187717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,16384,0.035608000225490995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,64,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,3584,0.01180266671710544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,3072,0.011398221883508893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,3072,0.011081777513027191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,2560,0.01034311122364468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,4096,0.02330844435426924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,3584,0.010739555789364709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,2048,0.0086986662613021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,2560,0.01666755477587382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,5120,0.013648889131016202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,2048,0.009766221874290043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,2048,0.014286221729384528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,10240,0.024681778417693243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,4096,0.011623111036088733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,1536,0.009355555805895064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,1536,0.012098666694429187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,768,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,3072,0.018967999352349173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,1024,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,512,0.004575110971927643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,768,0.008341333104504479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,2560,0.010512000156773461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,768,0.00869688888390859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,512,0.007977777885066139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,512,0.00756177802880605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,1536,0.007640000018808577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,128,0.003539555602603488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,256,0.006506666541099548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,1024,0.006001777946949005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,64,0.0031866667171319327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,128,0.0076657773719893555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,32,0.003195555466744635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,1024,0.00869688888390859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,64,0.006039111150635614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,65536,0.10285422537061904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,32,0.005992888990375731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,256,0.0038942222793896994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,51200,0.08640622430377537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,65536,0.12515821721818712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,16384,0.032666666640175715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,65536,0.26381511158413357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,256,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,12288,0.022201778160201177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,16384,0.034824000464545354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,128,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,16384,0.06765955686569214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,10240,0.01907999979125129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,12288,0.05186400148603651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,8192,0.013955555028385587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,10240,0.023970666858885024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,10240,0.04387022058169047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,7168,0.01314222233162986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,8192,0.017871111631393433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,8192,0.03609777821434869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,6144,0.011438222395049201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,51200,0.1004142231411404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,51200,0.20357067055172393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,7168,0.03241422110133701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,6144,0.015448000695970325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,4096,0.008667555948098501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,5120,0.013814222481515674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,5120,0.02436622149414486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,3584,0.007917332980367873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,4096,0.020523554748959012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,3584,0.02110311057832506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,3584,0.01868977811601427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,3072,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,7168,0.016568000117937725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,5120,0.01035288886891471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,3072,0.01681600014368693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,6144,0.02851822310023838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,12288,0.026865777042176988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,2048,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,2560,0.014852444330851236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,4096,0.012449777788586087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,2048,0.010048888623714447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,2048,0.012882666455374824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,1024,0.004575110971927643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,1536,0.00923822240697013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,1536,0.01109333336353302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,768,0.0042462220622433555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,1024,0.008664000365469191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,2560,0.00666933341158761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,512,0.003879111260175705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,768,0.00832799987660514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,768,0.008030222521887885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,2560,0.010699555277824402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,512,0.006960888703664144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,1536,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,128,0.003208000005947219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,256,0.007631111476156447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,256,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,3584,0.011759999725553723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,128,0.007351111206743453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,32,0.003195555466744635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,3072,0.011101333631409539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,64,0.005620444400442972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,65536,0.06621688604354858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,32,0.005628444254398346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,512,0.007677333222495184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,51200,0.05455199877421061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,65536,0.12555378013186985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,65536,0.2214346726735433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,51200,0.10020711024602254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,64,0.0028568889117903183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,12288,0.013456000222100152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,16384,0.05827377902136909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,1024,0.009030222064918941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,12288,0.02815377712249756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,12288,0.0446266664399041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,256,0.003208000005947219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,10240,0.025106666816605464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,8192,0.01388088862101237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,10240,0.03775022096104092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,16384,0.019184889064894784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,7168,0.011926222178671094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,8192,0.01830311119556427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,51200,0.17331466409895155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,7168,0.017768889665603638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,16384,0.035014222065607704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,128,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,5120,0.009359999663299983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,6144,0.016490666402710807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,4096,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,6144,0.024675556355052527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,5120,0.014159111513031853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,5120,0.02104088829623328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,4096,0.012410666379663678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,4096,0.01793688866827223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,3584,0.011809777882364062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,3584,0.016525333126386006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,8192,0.03084711233774821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,7168,0.027506666051016912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,3072,0.011447110937701331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,2560,0.00609688874748018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,3072,0.014831999937693277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,10240,0.014175110393100314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,2048,0.00553955551650789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,2560,0.010732444624106089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,1536,0.004948444250557157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,2048,0.010066666536860997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,2048,0.011559999651379056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,1024,0.004232888834344016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,1536,0.010085333552625444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,6144,0.010764444039927589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,768,0.003993777765168084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,1024,0.008728000024954477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,768,0.00848711116446389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,2560,0.013447999954223633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,768,0.007335111498832703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,1536,0.009389333426952362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,256,0.0032017777363459268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,512,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,3072,0.00664533343580034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,128,0.0029022221763928733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,256,0.007637333538797166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,64,0.002924444360865487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,128,0.007648000286685095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,32,0.0028408887899584244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,128,0.0052897776994440295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,65536,0.06365511152479383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,64,0.005249777601824866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,32,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,512,0.007958222594526079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,65536,0.21992088688744438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,16384,0.014478221535682678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,51200,0.17309867011176217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,256,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,10240,0.013119999733236102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,16384,0.05827644136216906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,8192,0.011424000064531961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,512,0.003633777714437909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,51200,0.05455999904208713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,8192,0.030547555949952867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,6144,0.009496000077989366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,7168,0.027091556125217017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,1024,0.008336000144481659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,6144,0.024375110864639282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,4096,0.007684444387753804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,12288,0.04457688993877835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,3584,0.007281777759393056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,7168,0.011057777537239922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,3072,0.006709333509206772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,3584,0.016454221473799813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,2560,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,5120,0.009343110852771336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,2048,0.0052382221652401825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,2560,0.013456888496875763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,1536,0.00490399988161193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,5120,0.02095555596881443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,2048,0.011437333292431302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,3584,0.007616000043021307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,768,0.003891555385457145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,10240,0.037271999650531344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,1536,0.010051555931568146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,1024,0.00835111074977451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,256,0.0031893334041039148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,768,0.007351111206743453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,128,0.0028737777223189673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,12288,0.014132445057233175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,64,0.0028702221396896574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,256,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,32,0.002943111169669363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,1024,0.004217777815130022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,65536,0.06467822525236341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,4096,0.01793599956565433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,32,0.0052826665341854095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,512,0.003567999849716822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,16384,0.013456000222100152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,3072,0.014869333969222175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,51200,0.17734311686621773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,12288,0.011739555332395764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,10240,0.010429333481523724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,128,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,8192,0.00904444439543618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,64,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,7168,0.008808000220192803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,10240,0.03753511110941569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,51200,0.053816000620524086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,8192,0.030472887886895075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,5120,0.008029333419269985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,7168,0.027519999278916255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,65536,0.22575643327501085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,6144,0.024303111765119765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,16384,0.058232002788119845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,5120,0.02104355560408698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,3584,0.008375110725561777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,3072,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,4096,0.017877333694034152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,12288,0.04480088750521342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,3584,0.01647644407219357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,2048,0.005239999956554837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,3072,0.014885332849290637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,6144,0.008496888809733922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,1536,0.004564444637960858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,2560,0.013108444710572561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,1024,0.0038746665749284956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,768,0.003887999802827835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,2048,0.01148000028398302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,512,0.006541333264774746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,1536,0.010080888867378235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,1024,0.008431110945012834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,768,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,128,0.0029048888633648553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,512,0.00657066661450598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,2560,0.007295110987292395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,128,0.005247999810510212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,32,0.0028799999919202593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,64,0.005161777966552311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,4096,0.009380444056457942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,256,0.003256888853179084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,64,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,512,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,256,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,32,0.005245333330498801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,16384,3.603732638888889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,16384,1.1030622058444552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,16384,3.9951519436306424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,51200,4.165099673800998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,12288,2.843272950914171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,12288,0.8190462324354383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,12288,2.875239054361979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,16384,3.418557273017036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,12288,2.526115629408095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,10240,0.7031546698676215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,10240,2.4274658626980252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,10240,2.365599102444119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,10240,2.1341768900553384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,8192,0.5675937864515517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,51200,11.614388359917534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,8192,1.9471698337131078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,8192,1.9316683875189886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,7168,0.5084017647637261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,8192,1.703495979309082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,7168,1.686796400282118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,7168,1.548904842800564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,6144,1.4780800077650282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,6144,0.43463733461168075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,7168,1.446105745103624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,5120,1.14301331837972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,6144,1.3395004272460938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,5120,0.36884087986416286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,5120,1.1120320426093209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,4096,0.9238408406575521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,5120,1.020517349243164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,4096,0.31139911545647514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,4096,0.8629235691494412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,6144,1.228250715467665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,3584,0.28535644213358563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,51200,16.43160163031684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,4096,0.795064025455051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,3584,0.7751875453525119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,3072,0.60971646838718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,3072,0.24421599176194933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,51200,11.681834750705294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,2560,0.5112755563524034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,3072,0.6562515364752876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,3584,0.8011022143893771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,2560,0.22736620903015137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,2560,0.5168755319383409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,2048,0.48102135128445095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,3584,0.6972808837890625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,2048,0.17956356207529703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,2048,0.43561598989698624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,3072,0.6278008884853786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,1536,0.14616799354553223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,1024,0.25221067004733616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,1536,0.33121776580810547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,1024,0.12532088491651747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,768,0.1770035558276706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,1024,0.2686142126719157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,2048,0.4532471232944065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,1024,0.2484133243560791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,512,0.13519644737243652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,1536,0.35309423340691465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,768,0.2027262184354994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,512,0.09114578035142686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,512,0.18376355701022676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,256,0.11172444290584987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,512,0.16519289546542698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,128,0.07865777942869398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,256,0.0808053347799513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,2560,0.5535288916693794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,256,0.16484622160593668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,256,0.13137688901689318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,128,0.07839733362197876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,768,0.21931821770138213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,128,0.10103644265068902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,32,0.07396088706122504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,1536,0.3635280132293701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,64,0.15745511319902208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,32,0.1599644422531128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,64,0.07093422280417548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,768,0.10354222191704644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,128,0.15518311659495035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,65536,4.334528181287977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,51200,3.3647973802354603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,16384,3.3001556396484375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,51200,8.524614969889322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,16384,3.327647103203667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,16384,0.8799209064907498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,65536,11.091451009114584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,12288,2.1972213321261935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,12288,0.6430444187588161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,16384,2.697217729356554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,12288,2.290326224433051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,12288,1.9623555077446833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,51200,9.161348130967882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,51200,12.738284640842013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,10240,1.9181058671739366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,10240,0.544782214694553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,10240,1.9031270345052083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,65536,15.943544175889757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,8192,0.4418035613165961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,10240,1.6541653739081488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,8192,1.4563075171576605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,7168,0.3879591094122992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,7168,1.3804693222045898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,8192,1.2909421920776367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,8192,1.4293013678656683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,6144,1.0526338153415256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,6144,0.34471466806199813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,7168,1.136606216430664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,5120,0.7654630872938367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,6144,1.0425519943237305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,5120,0.30216889911227757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,7168,1.191640853881836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,5120,0.8667582405938042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,4096,0.6218159993489584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,5120,0.7649848726060656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,4096,0.24509421984354654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,65536,12.777963426378038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,4096,0.668751981523302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,3584,0.2172373268339369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,4096,0.6284053590562608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,3072,0.4637502034505208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,3584,0.6028195487128364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,6144,0.9301697413126627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,3584,0.5461324585808648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,3072,0.5164088673061794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,3072,0.19441955619388154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,2560,0.3714737892150879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,2560,0.16914844512939453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,3072,0.47818930943806964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,3584,0.592741330464681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,2560,0.4339146614074707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,2048,0.14529155360327825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,2560,0.40468443764580625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,2048,0.333496888478597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,1024,0.19462845060560438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,1536,0.2701644367641873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,1024,0.09320533275604248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,2048,0.32798756493462455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,1024,0.2064675490061442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,768,0.16266932752397326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,1536,0.2300106684366862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,2048,0.35888089074028867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,512,0.11197688844468857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,768,0.16959555943806967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,768,0.16133333577050105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,256,0.08037866486443414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,512,0.07406310902701484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,512,0.13230844338734946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,1024,0.21466399563683403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,128,0.06135555770662096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,256,0.0654275549782647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,256,0.10487377643585205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,128,0.06300889121161567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,128,0.1227662165959676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,768,0.08181155390209623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,32,0.05367111166318258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,512,0.14520355065663657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,128,0.08104799853430854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,64,0.12441333134969075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,256,0.13041688336266413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,64,0.05211377806133694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,32,0.12623555130428737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,65536,1.2866995069715712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,1536,0.11650578180948894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,65536,3.5982517666286893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,1536,0.2817653285132514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,51200,0.9809324476453992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,51200,2.806172476874458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,16384,0.2814728948805067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,16384,0.9459280437893338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,16384,0.9624960157606336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,65536,4.9856588575575085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,12288,0.6848950915866427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,16384,0.8197911050584582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,51200,4.081967247856988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,12288,0.21976266966925728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,12288,0.6920391188727485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,10240,0.5839964548746744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,12288,0.6013893551296657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,10240,0.1925884485244751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,51200,2.921985838148329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,8192,0.501397344801161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,10240,0.5053617689344618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,8192,0.4684133529663086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,8192,0.1493448946211073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,65536,3.7545598347981772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,7168,0.44233067830403644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,8192,0.41265066464742023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,6144,0.32465865876939565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,7168,0.40357422828674316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,7168,0.35195467207166886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,6144,0.11647910541958278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,5120,0.27834399541219074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,6144,0.31307021776835126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,5120,0.10052888923221165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,5120,0.25647465387980145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,7168,0.13385955492655435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,4096,0.22541512383355033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,4096,0.08432800239986843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,3584,0.1950204504860772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,6144,0.4305493301815457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,3584,0.07595288753509521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,3584,0.1972959968778822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,3072,0.17517244815826416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,5120,0.28560177485148114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,3584,0.18484621577792695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,3072,0.06725333134333293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,3072,0.17696711752149794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,3072,0.1661484506395128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,2048,0.1197813351949056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,4096,0.20771822664472792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,2560,0.1385600037044949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,10240,0.6006346808539497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,2048,0.050776888926823936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,1536,0.09649155537287395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,2048,0.1160000032848782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,2560,0.13653155167897543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,1536,0.09734221961763169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,1024,0.06362399790022108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,1536,0.09383289019266765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,2560,0.05924533473120796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,768,0.048390222920311816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,1024,0.07418488793902926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,1024,0.0721893310546875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,2048,0.1223262283537123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,768,0.059788445631663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,768,0.060144888030158154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,1536,0.0420684450202518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,512,0.05166755451096428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,512,0.03006311257680257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,256,0.026226666238572862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,1024,0.03552000059021844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,256,0.04610310991605123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,128,0.022983999715911016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,512,0.04012177719010247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,128,0.043453332450654775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,128,0.03260711166593764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,512,0.049913777245415576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,64,0.02553066611289978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,32,0.02640177806218465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,256,0.02720977697107527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,64,0.0440577765305837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,256,0.03937866621547275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,32,0.044511109590530396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,768,0.03277777632077535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,2560,0.14758843845791286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,4096,0.21469510926140678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,128,0.026467555099063452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,51200,2.3172098795572915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,65536,2.6969138251410594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,65536,1.0083724127875435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,51200,0.7934204207526313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,16384,0.7231101989746094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,16384,0.7416986889309354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,16384,0.25180800755818683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,12288,0.5096462037828233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,65536,4.031859503851996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,65536,2.8791342841254344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,12288,0.17246222496032715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,10240,0.47813600964016384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,12288,0.4549564255608453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,51200,2.223891576131185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,10240,0.151711106300354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,51200,2.923954645792643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,16384,0.6438239945305718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,8192,0.3699653413560655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,8192,0.11696355872684056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,8192,0.3604319890340169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,8192,0.3096017837524414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,12288,0.5285706520080566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,7168,0.31887110074361164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,7168,0.10371910863452488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,7168,0.26843465699089897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,10240,0.46068265702989364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,6144,0.09169422255622016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,6144,0.261957327524821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,6144,0.23107645246717665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,7168,0.34059821234809023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,5120,0.0777368876669142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,5120,0.215719117058648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,4096,0.19069600105285645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,5120,0.19568178388807508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,3584,0.1338017781575521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,4096,0.17157777150472006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,4096,0.1605715619193183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,10240,0.38583911789788145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,5120,0.23457421196831596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,3072,0.13840444882710776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,3584,0.14444977707333034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,3072,0.05344177616967095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,3072,0.13368622461954752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,3072,0.125928004582723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,6144,0.2556240028805203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,2560,0.04682488905058967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,2048,0.09514933162265354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,2560,0.1087920003467136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,2048,0.04047022263209025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,1536,0.07167733377880521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,2048,0.09124888976414998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,2560,0.11248532931009929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,1536,0.07572799921035767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,1536,0.034063110748926796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,1024,0.048026667700873486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,1024,0.058132443163130015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,1024,0.057670222388373486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,768,0.04141244292259216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,1024,0.027890665663613215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,768,0.04587822159131368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,4096,0.06736355357699923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,768,0.04912977748446994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,512,0.030952887402640447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,768,0.02602488961484697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,512,0.02386222283045451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,3584,0.15289955668979222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,512,0.03885599970817566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,512,0.04047288828425937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,256,0.021189333664046392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,256,0.034663110971450806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,256,0.03232355581389533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,2560,0.11399111482832168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,128,0.03227644496493869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,128,0.02048711147573259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,128,0.026309332913822595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,2048,0.09507999817530315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,256,0.027610666222042505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,64,0.03261333372857835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,32,0.0326222214433882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,1536,0.07427289088567098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,64,0.02241955531968011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,32,0.023009777069091797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,128,0.01798488861984677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,65536,0.796421315934923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,65536,2.3185058169894748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,3584,0.05892622470855713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,51200,1.660602675543891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,51200,0.6475955645243326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,16384,0.18735288249121773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,16384,0.6005422274271647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,16384,0.5071688758002387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,65536,2.3660365210639105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,12288,0.4014844364590115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,16384,0.5888275570339626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,12288,0.1525582207573785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,10240,0.33837956852383083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,51200,1.8625147077772353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,12288,0.3736284573872884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,10240,0.12296088536580403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,65536,3.356686274210612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,10240,0.36976000997755265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,10240,0.3141404522789849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,8192,0.10063555505540635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,51200,2.704983181423611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,7168,0.2507395479414198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,8192,0.2590995629628499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,12288,0.4357911215888129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,6144,0.21318488650851777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,7168,0.25431021054585773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,7168,0.22858667373657227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,6144,0.08202311065461901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,6144,0.22078843911488852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,8192,0.2976800070868598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,6144,0.19339111116197374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,4096,0.14001333713531494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,5120,0.06832622157202826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,8192,0.30220267507765025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,5120,0.16610933674706355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,3584,0.11130755477481419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,4096,0.05710577964782715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,4096,0.14574489328596327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,7168,0.08743911319308811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,4096,0.13759644826253256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,3584,0.13089599874284533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,5120,0.17389155758751762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,3072,0.11350488662719727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,2560,0.0861733357111613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,5120,0.18299822012583414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,2560,0.09718488984637791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,2560,0.0915599995189243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,3072,0.10206222534179688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,3584,0.12005510595109727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,3072,0.044849776559405856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,2048,0.08176266484790377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,1536,0.05128622386190626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,2048,0.07795111338297527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,3584,0.05106844504674276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,1536,0.06354222032758924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,1024,0.036807109912236534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,1024,0.04900977677769131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,3072,0.10708800289365981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,1024,0.04917777909172905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,1024,0.023967999551031325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,768,0.03927910990185208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,2048,0.07351733578575982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,768,0.02146133283774058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,512,0.03209244542651706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,768,0.030224889516830444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,512,0.01982399986849891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,256,0.017255999975734286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,512,0.03494311041302151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,2560,0.04025777843263414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,512,0.02348088887002733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,128,0.015479111009173922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,256,0.027648889356189307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,2048,0.03457333313094245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,128,0.02713866697417365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,64,0.013198222551080914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,128,0.023668444818920557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,128,0.017786666750907898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,32,0.013411555853154926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,1536,0.06501155429416232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,64,0.027483556005689833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,32,0.026994667119450037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,256,0.028796444336573284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,65536,0.6613706482781304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,768,0.041486223538716636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,1536,0.028541333145565454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,256,0.01763911048571269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,65536,1.9181537628173828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,51200,1.6881537967258031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,16384,0.5116106669108073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,51200,0.5371146731906468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,16384,0.5031635496351454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,16384,0.4280986785888672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,65536,1.9467440711127386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,12288,0.3577591048346625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,12288,0.12039555443657769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,10240,0.2515973250071208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,12288,0.3134400049845378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,16384,0.15702577431996664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,10240,0.29483644167582196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,10240,0.10137066576215957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,8192,0.24936265415615508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,12288,0.38392088148328996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,10240,0.2602897750006782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,8192,0.08458044131596883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,8192,0.23654132419162324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,8192,0.2126613325542874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,7168,0.07526666588253446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,7168,0.20946933163536918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,65536,2.631084442138672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,6144,0.17822310659620497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,6144,0.06432266367806329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,6144,0.16155378023783365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,5120,0.14043554994795057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,7168,0.17875288592444527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,5120,0.05679111348258125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,5120,0.15026844872368708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,51200,1.470816930135091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,5120,0.13824000623491076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,3584,0.09994488954544067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,7168,0.19103288650512695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,4096,0.11785954899258083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,3584,0.1069564421971639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,3584,0.1029671099450853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,4096,0.10835199885898167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,51200,2.181886249118381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,3072,0.09373866849475437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,2560,0.07120444377263387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,3072,0.03706399930848016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,3072,0.09148799710803562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,2560,0.03315022256639268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,2560,0.08045244216918945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,2048,0.058896885977851025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,3584,0.04257333278656006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,2048,0.02924355533387926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,6144,0.17206311225891113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,1536,0.042118221521377563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,4096,0.04643999867969089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,1536,0.05351466602749295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,1536,0.023804444405767653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,1536,0.055757330523596875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,1024,0.029284444120195176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,2560,0.07822044690450032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,3072,0.07906933625539143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,768,0.025079111258188885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,1024,0.020678222179412842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,2048,0.06705511278576322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,1024,0.042456888490253024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,512,0.01832799944612715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,768,0.01862044466866387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,4096,0.12044000625610352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,512,0.016528000434239704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,512,0.027313777142100867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,1024,0.04103733433617486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,2048,0.06643555561701457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,768,0.03319199879964193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,128,0.014816888504558139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,256,0.015501333607567681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,768,0.03586488962173462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,512,0.029776000314288672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,128,0.022724444667498272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,64,0.01365066650840971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,128,0.019715555840068393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,256,0.02388711108101739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,64,0.02273155583275689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,32,0.022674666510687932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,256,0.02374222212367588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,256,0.014468444718254937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,32,0.013817778064144982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,65536,1.6011404461330836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,51200,1.296368916829427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,51200,0.4937884542677138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,65536,2.19887203640408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,65536,1.7061209148830836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,51200,1.9225618574354384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,16384,0.4581928782992893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,16384,0.1482871108584934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,16384,0.38177866405910915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,12288,0.2926551236046685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,12288,0.10686310794618394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,16384,0.4152488973405626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,10240,0.29894667201571995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,12288,0.2826684315999349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,10240,0.094832890563541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,10240,0.26272532675001353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,51200,1.336753739251031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,8192,0.20889066325293648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,8192,0.19171199533674452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,65536,0.6319608688354492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,7168,0.20250222418043348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,7168,0.06727644470002916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,7168,0.18365689118703207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,8192,0.2294835514492459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,6144,0.1787066724565294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,7168,0.169250660472446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,10240,0.2362062136332194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,5120,0.14237156179216173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,8192,0.07632088661193848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,6144,0.14742578400505915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,128,0.01569955547650655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,5120,0.05123733480771383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,4096,0.09033066696590847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,5120,0.1262248886956109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,4096,0.041430221663581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,3584,0.10011911392211914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,4096,0.10785155826144749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,4096,0.10501333077748616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,3584,0.03753066725201077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,3072,0.09120711353090073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,3584,0.09557866387897068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,6144,0.05925689140955607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,3584,0.09359911415312026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,6144,0.15869243939717612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,2560,0.07607199748357137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,3072,0.03391822179158529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,3072,0.08367644415961371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,5120,0.1335795587963528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,3072,0.08266399966345893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,2560,0.07187288999557495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,2560,0.07132889164818658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,2048,0.06012888749440511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,2048,0.06177510817845663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,1536,0.04832977718777127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,1536,0.02199999988079071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,1024,0.026897778113683064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,1536,0.049858666128582425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,1024,0.01830577850341797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,2048,0.059375113911098905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,768,0.021571555071406897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,1024,0.03879822293917338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,1536,0.03868177864286635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,768,0.01723022262255351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,768,0.029939555459552344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,768,0.032708442873424955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,512,0.02458222210407257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,256,0.014371555712487964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,12288,0.31601866086324054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,2560,0.030156443516413372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,256,0.02182666626241472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,2048,0.026699556244744196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,512,0.017641777793566387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,512,0.015653333730167814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,128,0.020635555187861126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,512,0.02786133355564541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,64,0.010041777458455827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,128,0.014496000276671516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,32,0.010432889064153036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,128,0.018579555882347953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,64,0.020560888780487906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,256,0.014476444986131458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,128,0.0119991112086508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,32,0.020280889338917203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,65536,0.5432568656073676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,1024,0.03708711266517639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,65536,1.4921991560194228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,256,0.021320000290870667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,51200,0.4120035436418321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,16384,0.35926222801208496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,16384,0.37558666865030926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,65536,1.556416935390896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,51200,1.435309304131402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,12288,0.2662542131212023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,51200,1.2497271431816948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,12288,0.09800889094670613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,51200,1.1792275110880535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,10240,0.2428204483456082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,12288,0.2574542098575168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,10240,0.08051377534866333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,8192,0.17253955205281576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,16384,0.12296177281273736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,10240,0.23515555593702528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,10240,0.22863109906514487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,7168,0.1740595499674479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,16384,0.3441297743055556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,8192,0.1843848890728421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,8192,0.17686843872070312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,6144,0.13445599873860678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,12288,0.28829601075914174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,7168,0.15914666652679443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,6144,0.1389591164059109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,6144,0.05211911267704434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,6144,0.1338906685511271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,8192,0.06571022007200453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,5120,0.11744888623555501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,4096,0.09522666533788045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,5120,0.11406400468614365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,7168,0.05817688835991753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,4096,0.03701688845952352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,4096,0.09464977847205268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,4096,0.09461600250667995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,65536,1.9209938049316406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,3584,0.03377155462900797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,3584,0.08369866344663833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,3072,0.06944088803397284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,5120,0.043624000416861645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,3072,0.030571556753582422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,3072,0.07378844420115153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,3072,0.07476444376839532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,7168,0.16187644004821777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,2048,0.04731111062897576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,2560,0.06359733475579156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,2560,0.0648346675766839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,5120,0.12302488750881618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,2048,0.024031110935741003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,1536,0.03934489025010003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,2048,0.05510044429037306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,2560,0.0544435547457801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,1536,0.04246044490072462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,1024,0.026583999395370483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,1536,0.04510400030348036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,3584,0.08571733368767633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,1024,0.017669333351982962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,1024,0.03251733382542928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,1024,0.035410665803485446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,2048,0.05311200022697449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,768,0.016632889707883198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,3584,0.08471733331680298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,1536,0.020364445116784837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,512,0.02186222208870782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,256,0.011600000162919363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,512,0.025431111454963684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,2560,0.02717688845263587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,256,0.019261333677503798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,256,0.013478221992651621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,256,0.019866666860050626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,768,0.02650755478276147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,768,0.030006223254733618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,128,0.018481777773963082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,128,0.017530666457282174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,32,0.010010666317409938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,512,0.014479999740918478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,64,0.01837688850031959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,768,0.021987555755509272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,32,0.018160000443458557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,128,0.010943999720944298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,512,0.016209777858522203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,128,0.014120888378885059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,65536,0.46194934844970703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,65536,1.2086888419257271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,51200,1.1174080106947157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,65536,1.7747831344604492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,65536,1.1826035181681316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,16384,0.38198044564988876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,51200,1.3111724853515625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,51200,0.9243102603488498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,16384,0.10827022128634983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,12288,0.2545368936326769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,16384,0.3169342146979438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,12288,0.08362666765848796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,16384,0.27483288447062176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,12288,0.20431110594007704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,10240,0.0694613324271308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,10240,0.19530667199028862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,8192,0.1514728864034017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,10240,0.17314310868581137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,8192,0.05615466833114624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,51200,0.35642488797505695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,8192,0.1426559951570299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,7168,0.05007466673851013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,64,0.009346666435400644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,7168,0.13815644052293566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,6144,0.11108799775441487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,6144,0.04359289010365804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,6144,0.12001155482398139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,8192,0.15733777152167425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,7168,0.13116799460517034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,5120,0.1012986699740092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,5120,0.09410399860805935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,5120,0.038358221451441445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,5120,0.09136533074908787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,4096,0.08122133546405368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,3584,0.06050133042865329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,4096,0.08008178075154622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,3584,0.07204977671305339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,3584,0.07163555754555596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,4096,0.07036888599395752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,6144,0.11038577556610107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,3072,0.06371466981040107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,10240,0.18564177883995903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,2560,0.04271911250220405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,3072,0.062900443871816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,7168,0.12543733914693198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,2560,0.054823110500971474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,3584,0.028575999869240656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,2048,0.03439199924468994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,2560,0.02386666668785943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,2048,0.02072266737620036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,1536,0.02698844505680932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,3072,0.026251556144820318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,1536,0.03705244594150119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,1536,0.03838133480813768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,1024,0.020025778147909377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,2560,0.05482044484880236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,2048,0.046077334218555026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,2048,0.04552533229192098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,768,0.016712889075279236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,1024,0.015263999501864115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,768,0.02350844442844391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,768,0.02516888909869724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,512,0.012784000072214337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,768,0.014038221703635322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,512,0.01926399932967292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,1024,0.028468443287743464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,1024,0.029504888587527808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,512,0.012853333519564735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,256,0.016920000314712524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,3072,0.052588442961374916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,256,0.018219555417696636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,256,0.012168000141779581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,128,0.016188444362746347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,12288,0.25764444139268666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,256,0.009689778089523315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,64,0.007826666865083907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,32,0.00776800016562144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,64,0.01626488897535536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,1536,0.01793333391348521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,128,0.008452444440788692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,4096,0.031818668047587075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,128,0.01236177815331353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,65536,1.115406248304579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,32,0.01587466730011834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,65536,1.365772459242079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,65536,0.9878319634331597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,128,0.01604977746804555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,16384,0.2693555619981554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,51200,0.3162622186872694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,51200,1.0217635896470811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,16384,0.10009333160188462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,51200,0.7546417978074816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,12288,0.18874755170610216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,16384,0.2416000101301405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,51200,0.851645310719808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,10240,0.17518666055467394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,12288,0.19904532697465685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,12288,0.18318400118086073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,10240,0.06398666567272611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,8192,0.12710489167107478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,10240,0.16255822446611193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,10240,0.14975910716586643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,8192,0.049386666880713574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,65536,0.4001946714189317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,7168,0.12476355499691433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,12288,0.08185955550935534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,7168,0.11642311678992377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,512,0.021408889028761122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,7168,0.10788177781634861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,6144,0.10001511043972439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,6144,0.09404444694519043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,8192,0.1313048866060045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,6144,0.037444445821974016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,5120,0.0850142240524292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,7168,0.0447324448161655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,5120,0.033260444800059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,16384,0.25848176744249135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,8192,0.12178844875759548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,4096,0.02832888894610935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,3584,0.0557182232538859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,6144,0.09605777925915188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,4096,0.06252266963322957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,3072,0.04386933313475715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,3584,0.06141422192255656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,3584,0.062163558271196157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,4096,0.06857777966393365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,3072,0.0540231102042728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,5120,0.07611111137602064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,3072,0.05401244428422716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,2560,0.04680266645219591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,2048,0.030050665140151978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,3584,0.025193777349260118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,2048,0.039493332306543984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,2560,0.0366906656159295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,2048,0.04029511080847846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,1536,0.025215999947653875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,2560,0.021035555336210463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,1536,0.031699554787741765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,2560,0.04690488841798571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,1536,0.0332897769080268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,1536,0.015875554747051664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,1024,0.013808888693650564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,1024,0.024476443727811176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,768,0.014464888307783337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,1024,0.02647466626432207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,2048,0.017855111095640395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,512,0.011839110818174152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,768,0.012784889174832238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,768,0.020167110694779288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,768,0.02243288854757945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,256,0.009031111167536842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,4096,0.06822755601671007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,512,0.01683111157682207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,512,0.018960000740157235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,128,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,256,0.014471999473041959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,256,0.016590222716331482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,128,0.013896889156765409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,3072,0.02296977738539378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,128,0.015524443652894763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,32,0.007663999994595845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,512,0.012058667010731168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,5120,0.08095466428332858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,32,0.013817778064144982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,1024,0.018159111340840656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,256,0.01166488892502255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,128,0.011289777855078379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,64,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,65536,0.9269857406616211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,51200,0.7614524629380969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,64,0.013981334037250943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,65536,1.1250311533610027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,16384,0.25966045591566295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,51200,0.3148239983452691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,65536,0.9598986307779948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,12288,0.18124889002905953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,16384,0.23140178786383736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,51200,0.7169617546929253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,65536,0.348674668206109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,16384,0.23595288064744738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,12288,0.18400000201331246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,10240,0.1416177749633789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,8192,0.11012799872292413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,10240,0.057224889596303306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,10240,0.14701955848269993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,10240,0.145632889535692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,8192,0.045124444696638316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,8192,0.11883378028869629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,51200,0.8484311103820801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,12288,0.06613066461351183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,7168,0.10509155856238471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,7168,0.10538311137093438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,12288,0.1812871164745755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,6144,0.035192890299691096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,6144,0.09078666898939346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,5120,0.07540977663464017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,6144,0.09224533372455174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,16384,0.10363199975755479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,4096,0.05320088730918037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,5120,0.07646577888064914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,5120,0.07928355534871419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,7168,0.04029422336154514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,6144,0.08384266826841567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,4096,0.026187555657492742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,3584,0.04916266600290934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,4096,0.06633066468768649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,3584,0.02427022159099579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,7168,0.09788978099822998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,3072,0.03829333186149597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,3584,0.06076266368230184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,5120,0.030827555391523574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,3072,0.048771556880739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,3072,0.052870220608181424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,4096,0.06212355693181356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,2560,0.04232977827390035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,2048,0.026222222381167944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,2560,0.019796444310082328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,8192,0.12069866392347549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,2048,0.03540088733037313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,1536,0.021185777253574792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,3072,0.022121777137120564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,1536,0.028745777077145044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,1024,0.015822221835454304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,2560,0.045992887682384916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,3584,0.05537333422236972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,1024,0.021673777037196692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,768,0.012839111189047495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,2048,0.039389332135518394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,2560,0.032518221272362605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,768,0.01254666679435306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,512,0.010286221901575724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,768,0.021560889151361253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,1536,0.03312622176276313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,512,0.015170666906568738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,512,0.011760888828171624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,256,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,1024,0.01312444441848331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,1024,0.02533600065443251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,256,0.011069333387745751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,256,0.015633778439627755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,768,0.01846577723821004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,128,0.012963555753231049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,1536,0.015028445257080926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,128,0.014462222655614218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,32,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,512,0.01861422260602315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,64,0.013006221916940479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,32,0.012776888906955719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,128,0.007291555404663086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,65536,0.7186675601535373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,128,0.011057777537239922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,65536,0.2999182277255588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,65536,1.003127098083496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,2048,0.01722400055991279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,51200,0.6440986527336968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,16384,0.1810844474368625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,51200,0.24209955003526476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,51200,0.7502880096435547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,51200,0.603133307562934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,16384,0.2433706654442681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,64,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,16384,0.19326310687594941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,10240,0.11644977993435329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,12288,0.15580088562435573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,12288,0.14692621760898167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,65536,0.7841706805759006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,256,0.013413333230548434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,10240,0.050141334533691406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,10240,0.12593777974446616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,12288,0.12344888846079509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,8192,0.1040266686015659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,7168,0.07645599709616767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,12288,0.058633777830335826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,7168,0.09255111217498779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,7168,0.09189955393473308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,6144,0.06674844688839383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,8192,0.09506666660308838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,6144,0.031204445494545832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,6144,0.08043644163343641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,6144,0.08189955684873793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,8192,0.10223199923833211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,5120,0.027500443988376196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,4096,0.05007644494374593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,7168,0.0348293317688836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,10240,0.13122844696044922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,4096,0.05619644456439548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,4096,0.023631999890009563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,16384,0.07513333029217191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,3584,0.05044177836842007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,5120,0.06819999880260892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,3584,0.05153510967890421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,3584,0.021570665968788996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,5120,0.07019022438261244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,3072,0.01995733380317688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,3072,0.044567998912599355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,2560,0.030730668041441176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,3072,0.04631911052597893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,5120,0.06000889009899563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,2048,0.02492977844344245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,2560,0.03875733415285746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,2560,0.040968888335757785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,2048,0.032518221272362605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,2048,0.015828443898095023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,1536,0.020644444558355544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,4096,0.057647109031677246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,8192,0.038935999075571694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,2560,0.017637333936161466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,1024,0.015660444895426433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,1536,0.028766221470303003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,3072,0.034274667501449585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,1024,0.019951999187469482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,768,0.01236977759334776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,1024,0.012487110992272695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,1024,0.02239022155602773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,3584,0.042146666182412036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,768,0.016673778494199116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,512,0.009684444301658207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,768,0.019574221637513902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,1536,0.014185777968830533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,256,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,512,0.01686755485004849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,2048,0.034565332863065935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,256,0.012254222399658628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,128,0.006667555620272954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,256,0.010405333505736457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,256,0.014705777168273926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,1536,0.026725333597924974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,64,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,128,0.011754666765530905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,128,0.012865777644846173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,64,0.011749332977665795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,32,0.011752888560295105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,65536,0.6462684207492405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,768,0.011745777395036487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,128,0.010372444159454769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,32,0.005921777751710679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,65536,0.28865332073635525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,512,0.010792888700962067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,65536,0.639950222439236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,51200,0.5298213428921169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,16384,0.15153600109948054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,51200,0.6012426482306586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,16384,0.07015199793709649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,51200,0.4861084620157878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,16384,0.1637680000729031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,12288,0.12465688917371963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,10240,0.11034844319025676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,12288,0.056681778695848256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,12288,0.13537600305345324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,512,0.014163555370436775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,10240,0.0483804444471995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,10240,0.11430399947696263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,51200,0.22505688667297363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,8192,0.03695733348528544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,16384,0.18149511019388834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,8192,0.09211644199159409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,7168,0.07845244142744276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,8192,0.08568266365263198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,6144,0.06488444407780965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,7168,0.033180445432662964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,65536,0.7760453224182129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,7168,0.08186133040322198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,5120,0.05371555685997009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,12288,0.12528088357713488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,6144,0.07128355238172743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,6144,0.06642399893866645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,10240,0.10273866520987616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,4096,0.04099199838108487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,5120,0.06006222301059299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,5120,0.0589324434598287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,3584,0.033385776811175875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,4096,0.04826133449872335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,7168,0.07703377803166707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,6144,0.029578665892283123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,8192,0.08418133523729111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,3584,0.04360266526540121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,3584,0.020460445019933913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,5120,0.026250667042202417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,3072,0.018773333893881906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,2560,0.02509511179394192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,3072,0.03832533293300205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,3072,0.03905155592494541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,4096,0.04792088932461209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,2048,0.02074399921629164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,2560,0.017234666479958426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,2560,0.03358133302794562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,2560,0.03393955694304572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,2048,0.015284443895022074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,2048,0.027986665566762287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,3584,0.04350044329961141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,2048,0.029425776667065088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,3072,0.028352889749738906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,1536,0.022882666852739122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,1536,0.024473778075642053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,768,0.010498666928874122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,1024,0.017479999197853934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,1024,0.019288000133302476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,4096,0.022387555903858606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,512,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,768,0.014813333749771118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,768,0.01684533390733931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,1536,0.01668711172209846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,512,0.012502222425407834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,1024,0.013090666797426013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,1536,0.012934221989578672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,512,0.01588888963063558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,1024,0.01184977756606208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,256,0.01110577748881446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,256,0.009986666341622671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,256,0.013633777697881063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,768,0.010857777463065254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,128,0.009905777871608734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,64,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,32,0.005893333504597346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,512,0.010489777558379702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,256,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,64,0.010382222632567087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,32,0.01074222226937612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,128,0.00591555568906996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,65536,0.532000011867947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,65536,0.2450204425387912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,65536,0.6421644422743056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,128,0.01037333326207267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,128,0.01330044451687071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,16384,0.13153155644734701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,51200,0.49911822213066953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,16384,0.06306844287448458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,12288,0.1019315587149726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,16384,0.15543556213378906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,12288,0.04718222220738729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,12288,0.11991911464267307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,51200,0.38720001114739316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,65536,0.6166115336947972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,51200,0.1899262269337972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,8192,0.06803733110427856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,10240,0.09751200013690525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,51200,0.49477243423461914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,16384,0.15393155150943333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,7168,0.05867822302712334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,8192,0.03136355678240458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,8192,0.07961066563924153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,8192,0.08396266566382514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,12288,0.12129688262939453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,6144,0.04478933413823446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,10240,0.08365422487258911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,7168,0.07419911358091566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,6144,0.02586222191651662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,6144,0.06127555502785576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,6144,0.06520089175966051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,10240,0.10144088665644328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,5120,0.022487110561794702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,5120,0.05606222152709961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,7168,0.028648889727062646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,4096,0.019290667441156175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,3584,0.030104888810051814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,10240,0.040976888603634305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,4096,0.03289066751797994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,3584,0.01796444422668881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,3072,0.023903111616770428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,4096,0.042456888490253024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,7168,0.0705804427464803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,3072,0.03382044368320041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,2560,0.020783111453056335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,3072,0.016545777519543965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,5120,0.05134044422043694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,2560,0.02924622098604838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,2048,0.01830222209294637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,2048,0.024824889169798955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,5120,0.03747555613517761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,3072,0.038266665405697294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,1536,0.014099554883109199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,2048,0.014149333039919535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,2048,0.02797422144148085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,3584,0.03817066550254822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,1024,0.010407110883129967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,1536,0.01241422196229299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,1536,0.02055288851261139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,1536,0.02379111117786831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,1024,0.011042666931947073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,1024,0.01575644479857551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,1024,0.018629334039158292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,4096,0.04739822281731499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,512,0.007066667079925537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,768,0.013447110851605734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,768,0.01657777859105004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,512,0.011405333048767514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,512,0.009711110757456886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,256,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,512,0.014186667071448432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,2560,0.015270221564504834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,128,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,256,0.00941066692272822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,256,0.010008888940016428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,256,0.013468444347381592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,64,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,128,0.009349333743254343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,32,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,128,0.011789333489206104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,64,0.009349333743254343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,32,0.009358222285906473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,2560,0.03335377905103896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,768,0.008611555728647444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,65536,0.18031111028459337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,128,0.009350222018029954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,65536,0.4637706544664171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,51200,0.2945324314965142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,65536,0.309350225660536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,51200,0.15005511707729763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,51200,0.41853777567545575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,3584,0.04235022266705831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,51200,0.3647608757019043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,16384,0.048672888014051646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,12288,0.0722915530204773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,16384,0.12116089132097031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,768,0.010469333165221743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,12288,0.10032177633709377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,10240,0.06029778056674533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,12288,0.09319377607769436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,10240,0.03376888897683885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,10240,0.0833466649055481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,8192,0.05057866705788506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,10240,0.07919555240207247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,16384,0.13152621852027044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,8192,0.06748800145255195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,7168,0.041640000210867986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,8192,0.06610221995247735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,12288,0.038188444243537054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,7168,0.02288088864750332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,7168,0.0595724450217353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,7168,0.060257779227362744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,5120,0.030769778622521296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,6144,0.02106399999724494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,6144,0.05173066589567396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,6144,0.051833778619766235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,4096,0.025099555651346844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,5120,0.044509334696663745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,5120,0.044842667049831815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,4096,0.03654044535424974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,3584,0.02272355556488037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,4096,0.03889155719015334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,16384,0.09787111149893866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,3584,0.033055110110176936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,3072,0.02032533288002014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,3584,0.03446400165557861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,3584,0.015021334091822306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,5120,0.018217777212460835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,8192,0.025978666212823655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,3072,0.014171555638313293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,3072,0.028920888900756836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,3072,0.031202667289310034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,4096,0.01633066601223416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,2048,0.0151582227812873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,2560,0.02539377742343479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,2560,0.026390223039521113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,1536,0.011955555942323474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,2048,0.011526222030321756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,2048,0.021874666213989258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,2048,0.023045332895384893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,1024,0.009770666559537252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,1536,0.01807466646035512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,1536,0.019707555572191875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,1024,0.014022222823566861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,1024,0.009815110928482479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,1024,0.015593777100245157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,2560,0.012784889174832238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,6144,0.03705777724583944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,512,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,768,0.009386666946940953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,768,0.014168888330459595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,512,0.010225777824719747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,512,0.009000000026490953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,256,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,2560,0.01769244505299462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,256,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,128,0.004957333207130432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,256,0.011071110765139261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,768,0.01203733351495531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,128,0.008355555435021719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,64,0.004604444321658877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,1536,0.01073688848151101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,32,0.004838222016890843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,512,0.012335110869672565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,65536,0.5287946595085992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,64,0.008380444513426887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,65536,0.2698337766859266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,32,0.008348444269763099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,51200,0.21286933951907686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,65536,0.17077155907948813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,65536,0.39622044563293457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,128,0.008628444539176093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,51200,0.13738311661614314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,16384,0.06733333402209811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,51200,0.31306934356689453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,51200,0.3536168999142117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,16384,0.04665866825315687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,12288,0.05313511027230156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,16384,0.11962666776445176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,12288,0.03600533472167121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,10240,0.04214844438764784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,128,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,10240,0.06474133332570393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,10240,0.07792266872194079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,10240,0.03211022085613675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,8192,0.03312089045842489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,12288,0.07751466830571492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,12288,0.09143644571304321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,8192,0.05280799998177422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,7168,0.028350220786200628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,8192,0.024175999893082514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,8192,0.0651715530289544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,7168,0.04662044511901008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,256,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,7168,0.05726399686601427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,6144,0.04081333345837063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,6144,0.05023733443684048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,6144,0.02012888921631707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,5120,0.022096888886557683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,768,0.00814222213294771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,5120,0.04349777764744229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,4096,0.01828177769978841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,65536,0.4520702362060547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,4096,0.029013332393434312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,3584,0.016753777861595154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,7168,0.022535110513369244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,3584,0.025927111506462097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,5120,0.03479289015134176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,3072,0.014720888601409065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,3584,0.03331111205948724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,3072,0.02323377794689602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,4096,0.015608888533380298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,2560,0.012429333395428128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,3072,0.013539555172125498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,16384,0.10228711366653442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,2560,0.020350222786267597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,5120,0.017722666263580322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,2560,0.025715554753939312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,3584,0.014478221535682678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,2048,0.011411555939250521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,2048,0.0173999998304579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,2048,0.022618666291236877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,4096,0.03681244452794393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,3072,0.029647111892700195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,1024,0.006959999601046245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,1536,0.010405333505736457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,6144,0.02500533395343357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,1536,0.01906222270594703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,768,0.006317333214812809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,1024,0.009704888694816166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,1024,0.011757333245542316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,1024,0.015211555692884656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,512,0.005325333525737126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,768,0.010088889135254754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,1536,0.0087333329849773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,768,0.014088888963063559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,256,0.004580444345871608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,512,0.008662222160233391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,512,0.012461333639091916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,256,0.0074417781498697065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,256,0.008431110945012834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,128,0.0042408886882993905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,128,0.007091555330488417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,128,0.010760889285140567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,128,0.008375110725561777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,512,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,32,0.003907555507289039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,1536,0.014520888527234396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,64,0.006999111009968652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,65536,0.20569066206614176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,32,0.007269333634111617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,768,0.009346666435400644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,65536,0.168649779425727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,65536,0.3451982339223226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,51200,0.1671466694937812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,65536,0.30704355239868164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,16384,0.055142223834991455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,51200,0.13752799563937718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,51200,0.2785031000773112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,51200,0.24088356229994032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,12288,0.041745778587129384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,16384,0.09035199880599976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,16384,0.08324622445636325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,12288,0.06823466883765326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,12288,0.06561422348022461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,10240,0.058232002788119845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,16384,0.04596977763705783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,10240,0.03164533442921109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,8192,0.025154666768179998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,64,0.003903111235962974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,8192,0.04703466759787666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,8192,0.04646311203638712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,8192,0.023522666758961145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,7168,0.021984888447655573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,2560,0.01242844429281023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,7168,0.02169955604606205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,7168,0.0417173339260949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,6144,0.01997599999109904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,7168,0.04173422190878126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,10240,0.03417511118782891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,5120,0.017091555727852713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,6144,0.020063999626371596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,6144,0.03670755690998501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,6144,0.036369777388042875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,5120,0.017440888616773818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,5120,0.03125155634350247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,4096,0.01456622282663981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,5120,0.03141155507829454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,256,0.010769777827792697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,3584,0.012940444052219391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,4096,0.01551999979548984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,4096,0.02589955594804552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,4096,0.02703022294574314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,3072,0.011914666328165265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,3584,0.02344266739156511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,3584,0.024393777052561443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,3072,0.020985777179400127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,2560,0.010654222634103564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,2048,0.011429333024554782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,2560,0.01869066721863217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,12288,0.036328001154793635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,3584,0.014546665880415173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,2560,0.019537778364287484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,3072,0.013460444079505073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,3072,0.021998221675554912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,1536,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,2048,0.011415999796655444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,10240,0.05587822198867798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,1536,0.013230221966902414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,1024,0.006504000061088138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,1536,0.014887111054526435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,2560,0.012466666599114736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,768,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,1024,0.009690666364298927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,1024,0.012423111332787408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,2048,0.01607377827167511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,2048,0.01761511133776771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,768,0.00907555553648207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,768,0.011239999698268043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,512,0.008678221868144141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,256,0.004224888980388641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,1536,0.010391111175219217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,2048,0.008817777865462834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,1024,0.01055111073785358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,256,0.008384000096056197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,128,0.0038720000949170855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,256,0.009393778112199571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,768,0.00941422250535753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,128,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,128,0.006640000061856375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,32,0.003863111138343811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,512,0.008038221961922115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,64,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,512,0.010453333457310995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,32,0.006640888750553131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,256,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,65536,0.1245573361714681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,65536,0.300385766559177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,51200,0.12391644053988987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,512,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,64,0.0035573335157500375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,16384,0.04275200102064344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,51200,0.10078489118152195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,51200,0.2423386573791504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,16384,0.03513333201408386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,16384,0.0808017783694797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,65536,0.15098488330841064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,12288,0.059960001044803195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,12288,0.06358489063051012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,65536,0.30142490069071454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,128,0.00904888908068339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,10240,0.050613333781560264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,8192,0.02072000006834666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,10240,0.024332443873087566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,10240,0.05475377705362108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,16384,0.08181955416997273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,7168,0.018735999862353008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,8192,0.04121955566936069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,12288,0.02767555581198798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,8192,0.0456008878019121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,10240,0.02548355526394314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,6144,0.016128889388508268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,7168,0.016580444243219163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,7168,0.03717155588997735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,5120,0.013945778210957846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,6144,0.015218666858143277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,6144,0.0325564444065094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,6144,0.03536622060669793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,4096,0.01180266671710544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,5120,0.027663111686706543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,5120,0.03119822343190511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,8192,0.018370666437678866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,4096,0.023244443866941664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,4096,0.012210666305489011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,4096,0.026738666825824316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,3584,0.011728888584507836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,3584,0.021014221840434607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,3072,0.009725333087974125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,3584,0.02408355474472046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,3072,0.01110577748881446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,5120,0.014021333720948962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,2560,0.009993777506881291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,3072,0.021792888641357422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,3584,0.011038222246699862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,2560,0.019131556153297424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,2048,0.00871022211180793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,7168,0.04042133357789781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,2048,0.014527999692493014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,51200,0.23644799656338164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,3072,0.018808888064490426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,1536,0.007679999702506595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,2048,0.016935111747847665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,2048,0.009835555321640438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,1024,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,1536,0.009367999931176504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,2560,0.010425777898894416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,1536,0.014818666709793938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,768,0.005374222166008419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,1024,0.012080888781282636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,768,0.008665777742862701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,12288,0.033109333780076765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,512,0.004561777744028303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,768,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,2560,0.01683288812637329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,512,0.010216000179449717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,512,0.007798222204049428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,256,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,768,0.011416888899273343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,128,0.0035386664999855887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,1536,0.012095111111799875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,1024,0.008667555948098501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,128,0.007689778175618913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,128,0.008678221868144141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,64,0.0032684444967243406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,32,0.0032426667296224167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,256,0.009378666679064432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,64,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,65536,0.10588533348507351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,32,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,512,0.007551111280918121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,51200,0.0858115553855896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,1024,0.009742221898502773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,65536,0.25845421685112846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,65536,0.2874408827887641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,16384,0.032292445500691734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,51200,0.20768000019921196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,51200,0.22572443220350477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,16384,0.06758844190173678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,16384,0.034790221187803484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,12288,0.025116443634033203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,16384,0.07842044697867499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,65536,0.12453956074184841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,128,0.006001777946949005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,12288,0.026660444007979497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,12288,0.05175911055670845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,12288,0.06070578098297119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,8192,0.017243555850452848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,10240,0.04366666740841336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,10240,0.05150844322310554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,256,0.007630222373538547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,7168,0.014466666513019137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,8192,0.03606133328543769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,8192,0.043344888422224254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,7168,0.032369779215918645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,7168,0.016320889194806416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,6144,0.01310755560795466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,10240,0.02409066590997908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,6144,0.027914666467242773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,5120,0.011781333221329583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,6144,0.015202666322390238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,8192,0.01800622211562263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,5120,0.024080889092551336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,5120,0.013616888887352414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,4096,0.008419555922349295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,256,0.003887999802827835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,10240,0.020462221569485135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,3584,0.008017777568764156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,4096,0.012331555287043253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,4096,0.025075554847717285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,6144,0.03381333417362637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,3584,0.018606222338146634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,3072,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,3584,0.022361778550677832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,3072,0.011071110765139261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,3072,0.016887111796273124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,2560,0.006676444576846228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,3072,0.020455999506844413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,7168,0.038543111748165555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,2048,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,4096,0.02050933407412635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,2560,0.014945778581831189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,51200,0.09939555327097575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,2560,0.017895999881956313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,1536,0.005557333429654439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,2048,0.009728888670603434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,2048,0.013050666285885705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,1024,0.004561777744028303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,1536,0.009400889277458191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,1536,0.011063111325105032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,1536,0.013980444934633044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,768,0.0041271112859249115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,1024,0.00906844437122345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,1024,0.01179377817445331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,2560,0.010481778118345471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,768,0.007985778152942657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,768,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,512,0.0038559999730851916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,768,0.01110222190618515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,3584,0.01146666705608368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,256,0.0035422220826148987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,512,0.009681777821646797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,1024,0.008661333057615492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,256,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,128,0.003208000005947219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,256,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,128,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,128,0.008360000120268928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,512,0.007807999849319458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,128,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,2048,0.01621955633163452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,32,0.0031991110493739447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,65536,0.06810577710469563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,64,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,32,0.005592888842026393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,256,0.008671110702885522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,51200,0.054619554016325206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,65536,0.12302754984961616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,65536,0.2847466733720568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,64,0.002899555489420891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,51200,0.17304000589582655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,51200,0.09923022323184544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,51200,0.2241448826260037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,12288,0.014142221874660917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,16384,0.05805155303743151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,16384,0.07748088571760389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,12288,0.04528266522619459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,12288,0.025705777936511572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,12288,0.05967999829186333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,512,0.0069848886794514125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,10240,0.03736889031198289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,8192,0.012138666378127204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,10240,0.05096266667048136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,5120,0.02907466557290819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,16384,0.03432710965474447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,7168,0.01145866678820716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,8192,0.04218666752179464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,10240,0.012816888590653738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,7168,0.027637332677841187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,7168,0.016885333591037326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,7168,0.03751555416319106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,10240,0.024313777685165405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,6144,0.02421244482199351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,5120,0.010081777969996134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,6144,0.015602666470739575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,8192,0.018292443619834054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,65536,0.22626754972669813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,5120,0.013931555880440606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,4096,0.008635555704434713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,5120,0.02880000074704488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,6144,0.011405333048767514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,3584,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,4096,0.012467555701732635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,16384,0.019526221685939364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,4096,0.024751999311976965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,3072,0.007299555672539606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,3584,0.011758222348160215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,3584,0.016531555189026725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,3584,0.022272888157102797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,3072,0.011071110765139261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,3072,0.014857777290874057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,3072,0.020317332612143624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,2560,0.010475555227862464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,2560,0.013456888496875763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,2560,0.017934223016103108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,4096,0.01807466646035512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,2048,0.010071111222108206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,1536,0.004953777624501122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,2048,0.015839111473825242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,8192,0.031013333135181006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,2560,0.006931555353932911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,1024,0.004210666649871402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,1536,0.010065777434243096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,1536,0.014072888427310519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,2048,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,1024,0.00833155545923445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,768,0.003931555483076307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,1024,0.011772444678677453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,768,0.00832444429397583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,768,0.007556444240940942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,768,0.010503110786279043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,6144,0.03279555506176419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,512,0.007872888611422645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,256,0.003261333331465721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,1024,0.008683555656009251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,256,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,2048,0.011744000017642975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,256,0.0075902218619982404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,256,0.008392000363932716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,128,0.005250666704442766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,512,0.009712888962692684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,64,0.0028782222006056043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,64,0.005221333354711533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,32,0.0032293332947625052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,512,0.003568888952334722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,32,0.0052782222628593445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,512,0.006659555352396435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,65536,0.06363377968470256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,51200,0.05095110999213325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,65536,0.22039911482069227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,16384,0.015428443749745687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,51200,0.17302489280700684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,128,0.003015111097031169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,12288,0.011747555600272285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,10240,0.010738666686746808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,16384,0.05822133355670505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,1536,0.009129777550697327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,10240,0.03751911057366265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,128,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,8192,0.03051911128891839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,6144,0.009009777671760982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,7168,0.027504889501465693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,5120,0.008608000146018134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,6144,0.0240639994541804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,4096,0.007648888561460707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,3584,0.007292444507280986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,7168,0.009367999931176504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,3072,0.006311999840868845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,3584,0.01657777859105004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,2560,0.005741333381997214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,3072,0.014868444866604276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,2048,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,2560,0.013292444248994192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,1536,0.004957333207130432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,5120,0.021048888564109802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,2048,0.011532444092962476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,1024,0.004214222232500712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,768,0.0038728887836138406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,1536,0.010078222387366826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,512,0.0034151110384199354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,1024,0.008302222523424361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,256,0.0032542221662071017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,768,0.007423111134105259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,128,0.0029146667155954572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,512,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,64,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,256,0.005586666779385672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,32,0.002879111096262932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,128,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,5120,0.020968000094095867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,65536,0.0662684440612793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,64,0.005239999956554837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,51200,0.05272622240914238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,32,0.005219555563396878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,65536,0.2269279956817627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,12288,0.010710222025712332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,16384,0.05834666887919108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,10240,0.009737778041097853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,12288,0.04485333296987745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,8192,0.009690666364298927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,10240,0.03746933407253689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,16384,0.01276000009642707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,8192,0.030769778622521296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,6144,0.008035555481910706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,7168,0.027426666683620874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,5120,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,6144,0.024296000599861145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,4096,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,12288,0.04416266745991177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,3584,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,4096,0.018013333280881245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,4096,0.01800622211562263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,3584,0.016491555505328707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,2560,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,3072,0.014859555496109856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,51200,0.1727688842349582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,2560,0.013202666408485837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,1536,0.004573333180612988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,8192,0.008970666262838576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,1024,0.003920000046491623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,7168,0.00831822223133511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,1536,0.010047111246320937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,1024,0.008355555435021719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,512,0.0031946665710873077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,768,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,2048,0.0052728888889153796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,512,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,128,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,5120,0.02106488909986284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,64,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,128,0.005274666680230035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,768,0.0038888889054457345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,64,0.005286222116814719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,32,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,256,0.0031973332580592898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,2048,0.011737777955002256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,3072,0.0069422221018208405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,32,0.0028657778683635923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,256,0.005658666706747479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,16384,1.1046311060587566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,16384,3.233885235256619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,12288,2.601940366956923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,16384,3.736194610595703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,12288,0.8009013070000542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,51200,4.072262234157987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,12288,2.772160847981771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,10240,0.6587235662672255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,10240,2.0534212324354386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,10240,2.1581929524739585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,8192,1.677443504333496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,8192,0.5558871163262261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,8192,1.7636017269558375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,7168,0.524280865987142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,7168,1.5690906312730577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,7168,1.4253146913316515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,6144,0.4272666772206624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,5120,1.0791511535644531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,5120,0.36824266115824383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,6144,1.2136560016208224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,51200,10.599378797743055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,5120,1.0055075751410592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,4096,0.3017057842678494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,4096,0.8775244288974338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,4096,0.8402329021030002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,6144,1.6165075302124023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,3584,0.6980195575290256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,3072,0.6510542233784994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,3072,0.23545866542392305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,3072,0.6102186838785807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,2560,0.517523553636339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,2560,0.2066533300611708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,3584,0.7554711235894097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,3584,0.2650986777411567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,2048,0.4417706595526801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,1536,0.37852801216973203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,2560,0.5277173254224989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,2048,0.4358728991614447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,1536,0.1416862275865343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,1536,0.33172355757819283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,1024,0.11325955390930176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,1024,0.24904444482591417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,1024,0.27530132399664986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,512,0.13824444346957737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,768,0.09976444641749065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,768,0.19878666930728486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,256,0.09453866879145305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,512,0.08921688795089722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,128,0.07896177636252509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,256,0.07975021998087566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,128,0.07698044512006971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,64,0.06837599807315402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,128,0.008340444829728868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,51200,15.13914828830295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,32,0.071779555744595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,2048,0.17461955547332764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,64,0.14738044473859999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,512,0.17130933867560494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,32,0.15740266111161974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,768,0.22243732876247832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,256,0.1534186601638794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,128,0.14458578162723118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,16384,2.4976613786485458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,51200,3.221237394544813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,16384,2.9097156524658203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,65536,4.062143961588542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,16384,0.8260035514831543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,12288,0.6422035429212782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,12288,2.273869408501519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,12288,2.0497618781195746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,10240,1.6034444173177083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,51200,7.931749979654948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,10240,0.5321253140767416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,8192,1.2287289301554363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,10240,1.7469600041707356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,8192,0.4288426770104303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,7168,0.39261600706312394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,8192,1.3235404756334093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,7168,1.2923644383748372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,65536,10.37503390842014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,6144,0.9828418095906576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,6144,0.33421601189507383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,7168,1.1332595613267686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,5120,0.9221697913275825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,6144,0.9622951083713107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,4096,0.6442409091525608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,4096,0.27798933453030056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,3584,0.5178409152560765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,5120,0.8047564294603137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,3584,0.2210035589006212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,3584,0.5724284383985732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,3072,0.45634041892157656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,51200,11.882249620225695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,3072,0.19155289067162407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,2560,0.3966551091935899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,3072,0.4804088804456923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,2560,0.16672889391581217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,4096,0.6366168657938639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,2048,0.3270684348212348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,2560,0.41766754786173504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,1536,0.2526480091942681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,1536,0.1150711112552219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,1536,0.26144178708394367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,5120,0.2890453338623047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,1024,0.09124266439014012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,1024,0.21554577350616455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,768,0.0795875522825453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,512,0.10244177447424994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,2048,0.13880356152852377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,1024,0.17770666546291777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,512,0.07185955842336018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,256,0.07039555576112536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,768,0.1322444412443373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,65536,15.006447686089409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,256,0.12112443976932102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,768,0.15671555201212564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,64,0.04974311259057787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,2048,0.3591022226545546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,512,0.135096894370185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,32,0.051833778619766235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,64,0.1165342198477851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,256,0.06435110833909777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,128,0.05853066841761271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,128,0.062012447251213916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,128,0.12792089250352648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,32,0.1174702246983846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,16384,0.8347039752536349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,51200,0.9564844767252604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,16384,0.28882132636176217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,16384,0.9041102727254232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,65536,3.232924355400933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,12288,0.6622977786593968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,65536,1.2772240108913846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,10240,0.5257404645284017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,10240,0.538299560546875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,51200,2.521536933051215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,8192,0.40676532851325137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,51200,3.9154328240288625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,12288,0.21888532903459337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,8192,0.1506968869103326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,65536,4.814945644802518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,7168,0.3508195612165663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,7168,0.13419466548495823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,8192,0.41797423362731934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,10240,0.18830755021837023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,6144,0.11348799864451091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,5120,0.2958773242102729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,12288,0.7254720264010959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,6144,0.3394640021853977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,5120,0.26303111182318795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,4096,0.08152088854047987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,3584,0.1634204387664795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,4096,0.21105510658688018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,3072,0.1440088881386651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,3584,0.07270577881071302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,3584,0.18393422497643364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,2560,0.1317217747370402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,3072,0.06456444660822551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,3072,0.16103022628360325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,2560,0.05716800027423435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,5120,0.10159644153383042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,4096,0.21269422107272676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,1536,0.08780799971686469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,2048,0.11342844698164199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,1536,0.04082755578888787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,1536,0.09074399868647258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,1024,0.03433600068092346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,6144,0.3098702165815565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,768,0.047285334931479565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,2048,0.10772622293896145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,512,0.03347200155258179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,768,0.031996445523367986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,2048,0.04941422078344557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,768,0.055606219503614635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,256,0.027638221780459087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,512,0.02950577934583028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,128,0.022610666023360357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,256,0.026867555247412786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,256,0.04315288861592611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,64,0.019174221489164565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,128,0.02569955587387085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,128,0.040884445110956825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,32,0.0200311111079322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,7168,0.38474133279588485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,64,0.04147466686036851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,2560,0.13731111420525446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,32,0.0414319998688168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,1024,0.056941330432891846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,512,0.04793333344989353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,1024,0.06880799929300944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,16384,0.5290657679239908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,65536,0.9913093778822156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,51200,0.7553466690911187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,16384,0.21350844701131186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,16384,0.664614200592041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,12288,0.5258382161458334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,12288,0.16442044576009116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,65536,2.675535202026367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,10240,0.41173421012030703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,10240,0.13979821734958225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,12288,0.7088044484456381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,8192,0.3516968886057536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,51200,2.7734247843424478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,7168,0.362610658009847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,8192,0.3437884383731418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,7168,0.10197422239515518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,65536,3.528281741672092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,7168,0.31796354717678493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,6144,0.24470755789015028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,6144,0.08962844477759467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,5120,0.18521511554718018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,4096,0.17380000485314262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,10240,0.43095021777682835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,5120,0.20189866754743788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,4096,0.06437600321239896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,3584,0.15732178423139784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,51200,2.045154571533203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,6144,0.24232623312208387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,3584,0.05841866466734144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,3072,0.11586399873097737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,3584,0.14330489105648464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,2560,0.10193866491317749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,3072,0.052073776721954346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,3072,0.12386843893263076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,2048,0.08984355794058906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,2560,0.045442667272355824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,2560,0.10658488670984904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,1536,0.06317155228720771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,2048,0.03928800092803107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,1024,0.05330400003327263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,1536,0.032405333386527166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,4096,0.1630693276723226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,768,0.040418667925728693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,1024,0.05331022209591336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,512,0.030079109801186457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,768,0.024076445235146418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,768,0.04252533449067009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,256,0.0211208893193139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,512,0.02201955517133077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,512,0.03537155522240533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,256,0.020103999310069613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,2048,0.08809066481060451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,256,0.03137066629197862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,64,0.01576266686121623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,1536,0.06998755534489949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,128,0.019932443896929424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,32,0.015484443969196744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,128,0.030376000536812678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,64,0.029988444513744775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,32,0.02993333339691162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,5120,0.07653599977493286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,128,0.018274666534529794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,1024,0.026692445079485577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,65536,0.7560888926188151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,8192,0.1128142277399699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,65536,1.9831617143419054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,51200,1.634804407755534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,51200,0.6086888843112522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,16384,0.18027733431922066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,12288,0.16875999503665498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,12288,0.37862666447957355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,10240,0.30723465813530815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,12288,0.47562668058607316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,10240,0.12009689542982314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,51200,2.3029999203152127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,8192,0.24481778674655488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,10240,0.3590719964769151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,8192,0.0979164441426595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,8192,0.2701999876234266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,7168,0.0855120023091634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,6144,0.19603378242916533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,16384,0.6058364444308811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,7168,0.24467733171251085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,5120,0.1769182284673055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,5120,0.0645608901977539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,5120,0.17253511481814912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,16384,0.5052142143249512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,4096,0.05395644572046068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,4096,0.13792533344692653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,3584,0.12112622790866429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,65536,3.0922906663682728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,3584,0.12193600336710612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,6144,0.07498044437832303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,2560,0.08247555626763238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,4096,0.1237128840552436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,2560,0.038330667548709445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,2560,0.09105866485171848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,7168,0.21839466359880236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,3072,0.10495733552508885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,1536,0.05375911129845513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,3072,0.04292800029118856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,6144,0.20461600356631807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,1536,0.06102399693595039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,1536,0.027560888065232172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,2048,0.07110044691297743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,1024,0.023289778166347083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,768,0.02875111169285244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,3584,0.04824622140990364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,768,0.02072533302836948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,2048,0.07547022236718072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,768,0.03654133280118307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,256,0.017465778523021273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,512,0.01906399925549825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,3072,0.10563466946283977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,128,0.014380445082982382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,256,0.01735555628935496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,256,0.02699377801683214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,128,0.017486222916179232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,128,0.0258586671617296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,64,0.012373333175977072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,32,0.012421333127551608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,2048,0.03303822212749057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,512,0.021335111724005804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,64,0.02552622225549486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,32,0.025398222936524287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,512,0.030417776770061914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,1024,0.038318223423428006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,1024,0.046335998508665294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,65536,0.65957244237264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,51200,0.5128489070468479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,16384,0.42501600583394367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,65536,1.7310720019870336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,16384,0.15067022376590306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,12288,0.35874578687879777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,51200,1.616958194308811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,16384,0.5404444270663792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,10240,0.2837119897206624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,12288,0.11615999539693196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,12288,0.3412791093190511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,10240,0.11995555294884576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,8192,0.2386524412367079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,51200,1.8907341427273219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,7168,0.0808933311038547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,6144,0.1752693388197157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,7168,0.21352177196078828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,6144,0.063573333952162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,5120,0.14378133085038927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,6144,0.1726213296254476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,5120,0.05386044581731161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,4096,0.11490933100382487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,5120,0.1422168943617079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,4096,0.04526933365397983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,8192,0.07983466651704577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,4096,0.1136888927883572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,8192,0.2286319997575548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,3584,0.04001244571473863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,3584,0.09944711128870647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,3072,0.03562488820817735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,2560,0.06891555256313749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,3072,0.0871466663148668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,2048,0.0510044429037306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,2560,0.03219466739230686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,2560,0.07485777801937528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,2048,0.027842667367723253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,3584,0.09818400277031793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,7168,0.18755911456214058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,1024,0.0314702226055993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,1536,0.023424888650576275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,3072,0.08118489053514269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,1536,0.05003644360436333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,768,0.024018666810459558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,1024,0.01978844404220581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,1024,0.03854488995340135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,768,0.01787644459141625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,768,0.031026668018764917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,256,0.015241778559154935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,512,0.0254737784465154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,1536,0.04244266615973579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,128,0.012418666647540199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,256,0.02254933284388648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,128,0.015488889482286243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,128,0.02165066699186961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,64,0.010375111467308467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,512,0.01882488860024346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,64,0.021645334031846788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,512,0.016359999775886536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,32,0.02149155570401086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,10240,0.2742062144809299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,32,0.011160000330872007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,65536,0.6052133242289225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,65536,1.6445751190185547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,51200,0.5195582177903917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,16384,0.413175106048584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,65536,2.316696802775065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,16384,0.1500071154700385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,16384,0.4208986759185791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,12288,0.3042977650960286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,65536,2.3751653035481772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,12288,0.10641066895590888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,10240,0.275819566514757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,10240,0.09085510836707221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,8192,0.22328355577256945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,10240,0.24586666954888237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,8192,0.08227644364039104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,7168,0.18266133467356363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,8192,0.19678221808539498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,51200,1.767044491238064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,7168,0.06386311186684503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,6144,0.17469422022501627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,7168,0.17365066210428873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,5120,0.11638932757907444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,6144,0.1486862235599094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,5120,0.04908000098334419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,4096,0.11079911390940349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,5120,0.12593421671125624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,3584,0.08550933334562515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,4096,0.040082666609022356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,4096,0.101218667295244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,2048,0.06252800093756782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,3584,0.036095112562179565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,3584,0.0890257755915324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,12288,0.3084613217247857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,3072,0.032200889454947576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,3072,0.078311112191942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,2560,0.029382222228580054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,2560,0.06688711378309461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,2048,0.05192177825503879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,2048,0.02544533378548092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,51200,1.2877235412597656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,1536,0.0392871101697286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,3072,0.070397330654992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,1024,0.028555555476082697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,1536,0.02160177793767717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,2560,0.05893066856596205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,1536,0.04479022158516777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,1024,0.018254222141371835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,1024,0.03452355662981669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,512,0.01702933344576094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,768,0.016639111770523917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,768,0.028059555424584284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,256,0.012757333616415659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,512,0.023016888234350417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,2048,0.05605066484875149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,128,0.011055111057228513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,256,0.020070221689012315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,64,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,128,0.014289778139856128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,128,0.01961866683430142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,32,0.010384889112578498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,64,0.01959022217326694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,32,0.019292443990707397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,512,0.015412444869677225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,256,0.015332443846596612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,256,0.014341332846217685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,768,0.022129777404997084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,65536,1.4720613691541884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,51200,0.3989866574605306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,16384,0.11841155423058404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,16384,0.4066133234235976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,6144,0.05587377813127306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,51200,1.367824872334798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,65536,0.5238746537102593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,12288,0.09132355451583862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,10240,0.19921955797407362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,51200,0.8712746832105848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,12288,0.25962665345933705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,8192,0.16154221693674722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,16384,0.327127112282647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,7168,0.14009066422780356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,8192,0.1720951133304172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,7168,0.056840889983707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,65536,1.8390560150146484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,12288,0.22587198681301543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,6144,0.13204888502756754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,7168,0.15145777331458196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,5120,0.09570844305886163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,6144,0.13034399350484213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,5120,0.042260444826549955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,10240,0.0781173308690389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,5120,0.11049333545896743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,10240,0.2129208909140693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,3584,0.07287110884984334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,4096,0.08834311034944321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,3072,0.06005777915318807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,3584,0.03267022305064731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,3072,0.029817776547537908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,3072,0.06941599978340997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,6144,0.05013155606057909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,2048,0.0446320010556115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,4096,0.07276266813278198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,2560,0.059212446212768555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,1536,0.03363200028737386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,2048,0.023580445183648005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,8192,0.06339822212855022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,2048,0.04950933323966133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,1024,0.025771554973390367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,1536,0.019820445113711886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,1536,0.039714667532179095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,3584,0.07868088616265191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,1024,0.017200888858901132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,2560,0.05481422278616163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,768,0.015541333291265698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,768,0.025055110454559326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,512,0.014149333039919535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,2560,0.027052443888452318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,512,0.020621332857343886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,128,0.010062221851613786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,256,0.01331022216214074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,256,0.018183110488785636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,64,0.009015999734401703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,128,0.01755555636352963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,768,0.02161688937081231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,512,0.015881778465376962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,64,0.017536888519922893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,32,0.017222222354676988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,256,0.01348622226052814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,4096,0.035754667388068304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,128,0.013803555733627744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,65536,0.44724533292982316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,32,0.008998221821255153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,1024,0.030439111921522353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,65536,1.3231564627753365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,16384,0.3015751043955485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,51200,0.38952000935872394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,65536,1.4498115115695531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,12288,0.22919999228583443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,51200,0.958738644917806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,16384,0.2951937781439887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,10240,0.18605332904391816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,10240,0.07035111056433783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,12288,0.21922043959299722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,10240,0.24351289537217882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,8192,0.055431110991372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,7168,0.12780000103844538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,8192,0.14641600184970432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,7168,0.05194311009513008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,7168,0.12963288360171848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,12288,0.08135288953781128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,6144,0.04320089022318522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,6144,0.11238933934105767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,4096,0.07151822249094646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,5120,0.09507733583450317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,4096,0.031381332212024264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,6144,0.10503821902804905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,3584,0.06769777668846978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,3584,0.02847111225128174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,3584,0.06773689058091906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,5120,0.08514133426878188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,5120,0.039232889811197914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,3072,0.02564711040920681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,3072,0.05986933575736152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,4096,0.07614133093092176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,2560,0.023306666149033442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,2048,0.03625333309173584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,2048,0.02035199933581882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,1536,0.028000887897279527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,3072,0.05811466773351034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,51200,1.1213422351413302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,1024,0.02030755579471588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,1536,0.017672888106769986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,2560,0.04424266682730781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,1536,0.0351804428630405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,768,0.016626665989557903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,1024,0.015061333775520325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,512,0.012345777617560493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,1024,0.026816000541051228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,768,0.013789333403110504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,16384,0.11716532707214355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,256,0.009711999860074785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,512,0.012776888906955719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,512,0.01819644371668498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,256,0.015951999359660678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,8192,0.1447920004526774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,128,0.012106666962305704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,64,0.015326221783955893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,32,0.00866844422287411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,2560,0.05167555477884081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,128,0.009019555317031013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,2048,0.04337866769896614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,64,0.007972444097201029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,65536,0.7203066613939074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,65536,0.39968087938096786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,768,0.022152000003390845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,51200,0.6925715340508355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,256,0.011971555650234222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,65536,1.1926924387613933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,51200,0.3478648927476671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,16384,0.09090844127866958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,128,0.01552977826860216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,51200,0.9242418077256945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,12288,0.06818311081992255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,10240,0.18072266048855254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,10240,0.06063466601901584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,10240,0.15260088443756104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,8192,0.04912889003753662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,32,0.015224888920783997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,7168,0.12083644337124294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,16384,0.2740844355689155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,7168,0.04361600014898512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,7168,0.108824888865153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,12288,0.19998755719926622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,6144,0.10432533423105876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,5120,0.08362044228447808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,6144,0.03838933176464505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,6144,0.09420622057384914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,4096,0.06604799959394667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,5120,0.07877333296669854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,16384,0.21663644578721789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,8192,0.12185511324140762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,4096,0.06392977635065715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,3584,0.02459466622935401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,3584,0.05676533116234673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,3072,0.02216888964176178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,3072,0.05064710974693298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,8192,0.13770311408572727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,5120,0.031493332650926374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,2560,0.020056888461112976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,2048,0.03319822086228265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,4096,0.02771111163828108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,12288,0.16287022166781956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,2048,0.017476444443066914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,2048,0.03678399986690945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,1536,0.01549600064754486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,1536,0.029339555237028334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,1024,0.018388445178667705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,2560,0.036947555012173124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,768,0.014862222803963555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,1024,0.013542222479979197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,1024,0.022437334060668945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,3584,0.061272886064317494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,512,0.011412444214026133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,768,0.019261333677503798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,256,0.008795555267069075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,512,0.011970666547616323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,512,0.01590399940808614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,256,0.01167555567291048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,256,0.013803555733627744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,128,0.01109955542617374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,128,0.013455111119482251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,64,0.00683822234471639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,32,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,2560,0.04364000095261467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,768,0.012735111018021902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,64,0.013456000222100152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,1536,0.024243555135197107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,32,0.01330133361948861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,128,0.007280000381999546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,65536,0.6594017876519097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,51200,0.2728239960140652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,51200,0.6219742033216689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,16384,0.08232622014151679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,12288,0.15581866105397543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,65536,1.050506697760688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,16384,0.22722487979465059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,12288,0.06666577524609037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,12288,0.1625759998957316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,10240,0.057143999470604785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,8192,0.11035111215379502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,65536,0.33975201182895237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,16384,0.21804800298478869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,7168,0.0952862236234877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,8192,0.10946310891045465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,7168,0.03929600119590759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,51200,0.7782470915052625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,10240,0.13137511412302652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,6144,0.03552000059021844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,6144,0.08500444226794773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,10240,0.13514667087131077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,5120,0.030699554416868422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,4096,0.05203466614087423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,5120,0.07124888896942139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,8192,0.044365333186255566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,3584,0.04257955484920078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,4096,0.026103999879625108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,3072,0.036700444089041814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,6144,0.08023555411232842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,3584,0.052058666944503784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,2560,0.029676444000667993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,3072,0.05064533485306633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,5120,0.05892088678148058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,2560,0.019706666469573975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,2560,0.03997333182228936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,2048,0.017306667235162523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,1536,0.020655110478401184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,4096,0.058223108450571694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,3584,0.02409066590997908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,7168,0.0971368882391188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,1536,0.014552889598740472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,3072,0.021716444028748408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,768,0.012235555383894177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,1024,0.013227555486891003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,1024,0.02126755482620663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,2048,0.027621333797772724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,512,0.009724444813198512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,768,0.012423999607563019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,768,0.017628444565667045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,512,0.011580444044537015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,512,0.014532445205582513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,256,0.01094044413831499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,256,0.012798222402731577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,1024,0.015494222442309061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,1536,0.027449778384632532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,3072,0.04626577761438158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,64,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,64,0.012320888539155325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,256,0.00755555596616533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,32,0.012280888855457306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,2048,0.033871110942628645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,128,0.007017778025733099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,65536,0.2921004560258654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,51200,0.42878044976128477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,128,0.010778666370444827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,16384,0.15304088592529297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,51200,0.23106045193142363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,65536,0.9255715476142036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,12288,0.1226746638615926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,16384,0.07238844368192884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,51200,0.7041404512193469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,16384,0.19249155786302355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,12288,0.05606933434804281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,12288,0.14987733629014757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,65536,0.6928248935275607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,10240,0.12038311693403457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,128,0.012431110772821637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,8192,0.03762400150299072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,7168,0.07677777608235677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,7168,0.034450666771994695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,7168,0.08684355682796902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,6144,0.06504088640213013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,10240,0.10501511229409112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,5120,0.04953333404329088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,10240,0.049060444037119545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,6144,0.07505866554048327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,32,0.00666133314371109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,4096,0.04198044538497925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,5120,0.06417510906855266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,3584,0.037668443388409086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,4096,0.05226222342915005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,3584,0.02106222179200914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,8192,0.08763021892971462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,3072,0.04184800055291918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,8192,0.09786755508846706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,2560,0.027437332603666518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,4096,0.02308622168170081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,2048,0.023020444644822016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,2560,0.03612800108061896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,2048,0.015609777635998197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,1536,0.01903199983967675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,5120,0.026927999324268762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,1536,0.013955555028385587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,1024,0.01440533333354526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,2560,0.017455110947291057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,1024,0.018751111295488145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,768,0.011771555576059552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,3584,0.047064890464146934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,768,0.011402666568756104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,768,0.015857777661747403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,3072,0.018989332848125033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,512,0.013365333278973898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,512,0.010710222025712332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,256,0.007177777588367462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,1024,0.012135999898115793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,256,0.01016266644001007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,128,0.006325333482689328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,256,0.011745777395036487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,64,0.00591466658645206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,128,0.010381333529949188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,32,0.00591822216908137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,512,0.010054222411579555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,64,0.011311111350854238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,1536,0.024367999699380662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,32,0.011149333582984077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,65536,0.5358835326300727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,65536,0.26952889230516225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,51200,0.39600356419881183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,6144,0.030585779084099665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,16384,0.14168177710639104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,51200,0.21402311325073242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,65536,0.7402044402228461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,12288,0.11199377642737494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,16384,0.06942133108774821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,16384,0.16961067252688936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,10240,0.09514311287138198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,12288,0.05407199925846524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,12288,0.12933866182963052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,8192,0.07262221972147624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,10240,0.04717155628734165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,10240,0.10746666457917954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,8192,0.03641422258483039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,3072,0.03181688984235128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,7168,0.03276088833808899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,6144,0.05275288886494107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,7168,0.07731822464201185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,5120,0.04939111073811849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,51200,0.5778320100572374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,5120,0.0255404445860121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,4096,0.036715557177861534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,7168,0.06528444422615899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,2048,0.03087377879354689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,4096,0.022031999296612207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,3584,0.031791110833485924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,128,0.011355555719799466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,3072,0.027015111512607996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,3584,0.04183377822240194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,6144,0.06714489062627156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,2560,0.02459288967980279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,3072,0.01846577723821004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,3072,0.037137776613235474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,2048,0.020364445116784837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,2560,0.016876444220542908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,2560,0.03196177879969279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,1536,0.015881778465376962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,2048,0.015271110667122735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,4096,0.04605244596799215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,1024,0.01166044423977534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,3584,0.020064888728989493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,1024,0.016865778300497267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,5120,0.05623822079764473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,1024,0.011696000066068439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,8192,0.08726221985287136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,768,0.014226666755146451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,512,0.007718222008811102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,768,0.010801778071456485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,6144,0.029892444610595703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,256,0.00665155549844106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,512,0.010341333018408881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,512,0.011818666425016193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,256,0.00980533328321245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,256,0.01073066641887029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,128,0.009780444204807281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,128,0.010050666828950247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,64,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,32,0.005492444253630108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,1536,0.012995555169052549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,1536,0.022055110997623865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,32,0.01015733347998725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,65536,0.36386577288309735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,128,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,768,0.009731555150614845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,65536,0.22911643981933594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,51200,0.28526398870680064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,16384,0.11512622568342422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,51200,0.18576533264584014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,12288,0.0856506692038642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,2048,0.02698933415942722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,51200,0.45683908462524414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,12288,0.0454293323887719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,64,0.010048888623714447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,12288,0.10776266786787246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,8192,0.05718311336305407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,10240,0.03990844554371304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,10240,0.08966933356391059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,8192,0.031144890520307753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,8192,0.07369866636064318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,7168,0.02794577678044637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,7168,0.06477244695027669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,65536,0.578205320570204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,5120,0.033972445461485125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,6144,0.02502577834659153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,16384,0.14214311705695257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,6144,0.0565226674079895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,4096,0.028637333048714533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,5120,0.02238133384121789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,5120,0.047424889273113675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,3584,0.025088000628683303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,7168,0.05307999915546841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,4096,0.03918577896224128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,3072,0.023429332507981196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,3584,0.01759288873937395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,3584,0.035234666532940336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,2560,0.019370666808552213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,3072,0.016196444630622864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,2048,0.016957332690556843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,2560,0.015014222926563688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,16384,0.06520889202753703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,2560,0.026994667119450037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,1536,0.01402844488620758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,2048,0.013911111487282647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,10240,0.0739004413286845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,1024,0.010245333115259806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,1536,0.01218933363755544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,768,0.00869599978129069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,4096,0.019163555569118924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,1024,0.01465777721669939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,512,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,768,0.012554667062229581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,6144,0.042544888125525586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,256,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,512,0.009727111293209923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,512,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,128,0.005140444470776452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,256,0.009353777600659264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,256,0.009413333402739631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,128,0.009389333426952362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,1536,0.019093333019150626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,1024,0.01091288857989841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,768,0.01038933296998342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,32,0.009107555780145857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,3072,0.031116445859273274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,65536,0.34044978353712296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,65536,0.17387466960483125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,64,0.004565333326657613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,128,0.009010666774378883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,32,0.004561777744028303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,16384,0.0898711085319519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,51200,0.14105244477589926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,12288,0.06634844674004449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,16384,0.12004000610775417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,12288,0.03620533479584588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,12288,0.0903315544128418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,10240,0.056227558188968234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,51200,0.24817956818474662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,8192,0.04208888941340976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,65536,0.4692595799763997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,10240,0.07561688952975802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,7168,0.03778666588995192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,8192,0.06176977687411838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,16384,0.04681866698794895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,7168,0.022837332553333704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,6144,0.03285600079430474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,2048,0.022964444425370958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,10240,0.03243822190496657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,5120,0.027841776609420776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,8192,0.02509511179394192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,51200,0.37886399692959255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,5120,0.018192888961897958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,4096,0.02260444396071964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,5120,0.040532443258497454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,3584,0.019688889384269714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,4096,0.015828443898095023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,4096,0.03336800138155619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,6144,0.04770311050944858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,3072,0.018124444617165458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,3584,0.030226667722066242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,2560,0.016000888413853116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,3072,0.013787555197874704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,3072,0.026557332939571802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,2048,0.013210666676362356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,2560,0.023286221755875483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,1536,0.011062222222487131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,7168,0.05452444487147861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,2048,0.011529777612951068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,2048,0.01978222197956509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,1536,0.010743111371994019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,1536,0.016680889659457736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,6144,0.020405333903100755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,768,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,1024,0.013009777499569787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,64,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,768,0.009367999931176504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,768,0.011061333119869232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,512,0.008828444613350762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,256,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,512,0.009701333112186855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,128,0.004601777841647466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,256,0.008454221818182204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,128,0.00850577818022834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,3584,0.01481066644191742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,1024,0.009694221946928237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,32,0.004548444516128964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,512,0.0064479998416370815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,64,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,2560,0.012685333689053854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,32,0.008028444316652086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,256,0.00849777791235182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,51200,0.18863111072116426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,65536,0.3811262183719211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,128,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,64,0.0042364444169733255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,16384,0.06388622522354126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,12288,0.04812888966666328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,51200,0.29681422975328237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,16384,0.04441155658827888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,16384,0.096888886557685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,12288,0.03489244315359328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,65536,0.241559108098348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,12288,0.07364444600211249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,10240,0.031531555785073176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,10240,0.061362663904825844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,1024,0.00867733359336853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,8192,0.02374222212367588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,8192,0.050047109524408974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,6144,0.02141155633661482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,7168,0.021760889225535925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,7168,0.044418666097852916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,5120,0.019069333871205647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,10240,0.03823733329772949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,8192,0.02776355544726054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,4096,0.015927111109097798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,65536,0.1660782231224908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,7168,0.024635555015669927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,3584,0.014516444669829475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,4096,0.01531377765867445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,51200,0.13168266084459093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,3072,0.01277511070171992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,3584,0.014356444279352823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,6144,0.019766221443812054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,3584,0.02511288887924618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,2560,0.01088444474670622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,6144,0.03898933198716905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,2048,0.009938666390048133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,2560,0.012312889099121094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,5120,0.03339733348952399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,2560,0.01963644391960568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,2048,0.011126221881972419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,2048,0.016957332690556843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,1024,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,1536,0.01036266651418474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,1536,0.014175999495718213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,768,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,1024,0.009681777821646797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,1024,0.011399110986126794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,512,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,3072,0.013493333425786761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,3072,0.02238222294383579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,256,0.004583999928500918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,512,0.00869599978129069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,1536,0.008013333711359236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,512,0.008388444781303406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,128,0.0038977778620190094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,256,0.007575111256705389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,4096,0.027936001618703205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,64,0.0038986665507157645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,128,0.008349333372380998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,128,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,768,0.00907555553648207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,768,0.009729777773221334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,65536,0.21418133046891955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,256,0.008360889222886827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,65536,0.32671022415161133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,51200,0.1589288844002618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,32,0.0037342223028341928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,51200,0.12889066007402208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,64,0.0069982219073507525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,5120,0.017289777596791584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,16384,0.04398666818936666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,16384,0.08575466606352065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,65536,0.16940977838304308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,10240,0.030637333790461223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,12288,0.03444622291458978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,12288,0.06471733252207439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,8192,0.02532177832391527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,10240,0.03094666533999973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,7168,0.02238311039076911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,8192,0.023132445083724126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,16384,0.054047111007902354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,8192,0.044446223311954074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,6144,0.019508444600635104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,7168,0.02141600019401974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,5120,0.016755556066830952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,6144,0.019374221563339233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,12288,0.039000001218583845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,4096,0.013968000809351603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,5120,0.017315554949972365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,5120,0.029725333054860432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,3584,0.011307555768224927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,4096,0.025077333052953083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,10240,0.05463911096254984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,3584,0.014136888914638095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,3584,0.022426666484938726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,51200,0.2565831078423394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,2560,0.008936000367005667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,3072,0.013129777378506131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,3072,0.02032888929049174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,2048,0.00814577771557702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,2560,0.012100444071822695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,2560,0.017907554904619854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,1536,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,2048,0.01107022249036365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,4096,0.015328889091809591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,2048,0.0155102229780621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,1536,0.01036800030204985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,1536,0.012987555729018318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,3072,0.010430222584141625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,1024,0.009359999663299983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,1024,0.010384000009960597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,512,0.004563555535342959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,32,0.007070221834712558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,768,0.00902133352226681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,256,0.0038693332009845306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,512,0.008692444198661381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,512,0.007717333734035492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,128,0.003887999802827835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,256,0.008350222474998897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,256,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,64,0.003556444413132138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,128,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,128,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,768,0.005231110999981563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,7168,0.040031111902660794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,65536,0.17019377814398873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,6144,0.0350195566813151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,51200,0.13222400347391763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,1024,0.0057502223385704895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,32,0.0038773334688610504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,51200,0.09627111090554132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,16384,0.03482666611671448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,768,0.00904266701804267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,12288,0.028852442900339764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,65536,0.11967021889156765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,12288,0.056629333231184215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,12288,0.025889777474933203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,51200,0.2247555520799425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,10240,0.04805955621931288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,64,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,16384,0.07453155517578125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,8192,0.01772622267405192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,8192,0.03938577903641595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,7168,0.015048000547620984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,65536,0.284351110458374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,10240,0.02146133283774058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,7168,0.01640266676743825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,7168,0.03501333461867438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,5120,0.011376000112957425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,6144,0.0307644440068139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,32,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,5120,0.013451555536852943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,4096,0.010388444695207808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,4096,0.022239110536045496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,4096,0.011763555308183035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,6144,0.014871999621391296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,16384,0.03303733468055725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,3584,0.020239111449983384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,3584,0.011414222419261932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,3072,0.008685333033402761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,3072,0.010875555376211802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,2560,0.007322666545708974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,3584,0.0094355551732911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,2560,0.01609511176745097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,2048,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,2560,0.01038577738735411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,5120,0.02663999961482154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,2048,0.009720000127951304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,2048,0.013842666314707862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,1024,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,1536,0.009240888886981541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,1536,0.011607999602953592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,768,0.004613333278232151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,1024,0.00868088917599784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,1024,0.009402666654851701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,512,0.003916444463862313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,768,0.008307555483447181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,8192,0.017014222012625802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,256,0.0038453332251972626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,512,0.008000888758235509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,512,0.007289778027269576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,128,0.0032302221904198327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,256,0.006551110910044775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,6144,0.013770666387346057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,128,0.007400000260935889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,128,0.005976888868543837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,32,0.003579555700222651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,64,0.005990222096443176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,10240,0.023227555884255305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,32,0.005998222364319696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,1536,0.005980444451173146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,256,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,51200,0.06796266635258992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,65536,0.2431253327263726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,3072,0.01819022165404426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,16384,0.024772443705134924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,51200,0.19125689400566948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,12288,0.01921688848071628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,16384,0.06451200114356147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,65536,0.0843271083301968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,10240,0.019091554813914828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,12288,0.025024000141355727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,12288,0.048723555273479886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,65536,0.11871466371748184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,8192,0.016007110476493835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,10240,0.04121422105365329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,51200,0.09550844298468696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,7168,0.014499555031458536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,8192,0.03402222196261088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,6144,0.013020444247457715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,7168,0.0162453336848153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,7168,0.030243555704752605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,768,0.008310221963458592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,6144,0.014723555909262763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,6144,0.026556443836953905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,4096,0.011065777805116443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,5120,0.0229759994480345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,10240,0.02278222143650055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,3584,0.008028444316652086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,4096,0.011767999993430244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,64,0.0032213332338465583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,8192,0.01755911111831665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,3072,0.006959999601046245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,3584,0.011336888704035016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,2560,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,3584,0.017603556315104168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,3072,0.010718222293588849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,16384,0.032467554012934365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,5120,0.011423110961914062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,2560,0.014144889182514615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,1536,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,2048,0.012485333614879184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,1536,0.009205333060688442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,1536,0.010482666393121084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,4096,0.01962844365172916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,1024,0.00868088917599784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,768,0.0041786668201287585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,2048,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,2560,0.010374222364690568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,512,0.0038844446341196695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,768,0.008217777642938826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,768,0.007808888951937358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,256,0.003256888853179084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,512,0.007816000117195977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,512,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,256,0.007345777418878343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,256,0.005979555762476391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,128,0.007619555625650618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,1024,0.008692444198661381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,3072,0.016075554821226332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,32,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,5120,0.013491555220550962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,128,0.0031884445084465873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,65536,0.06144622299406263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,128,0.005636444522274866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,65536,0.20972800254821777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,65536,0.11767821841769749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,2048,0.009716444545321995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,51200,0.0951111117998759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,32,0.005628444254398346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,16384,0.015782222151756287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,51200,0.16099733776516384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,12288,0.01348444405529234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,16384,0.032109333409203425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,64,0.003160888950030009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,12288,0.02530755599339803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,12288,0.04177511069509718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,10240,0.023717333873112995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,64,0.005626666463083691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,10240,0.03501244386037191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,7168,0.012480888929631976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,8192,0.017234666479958426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,8192,0.029091556866963703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,6144,0.012131555212868584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,7168,0.0258879992696974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,16384,0.05424177646636963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,6144,0.015199111567603217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,5120,0.010079999764760336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,8192,0.011060444845093621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,5120,0.013462222284740873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,4096,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,5120,0.02010222276051839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,3584,0.00869511150651508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,4096,0.012071111136012606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,1024,0.004555555681387584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,4096,0.017136888371573556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,7168,0.015875554747051664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,3584,0.015704888436529372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,3072,0.01073066641887029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,10240,0.01221244451072481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,6144,0.023017777336968318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,2048,0.005578666511509154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,2560,0.010384000009960597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,2560,0.012764444781674279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,51200,0.05060266786151462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,1536,0.00490044429898262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,2048,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,1024,0.004230222354332606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,1536,0.0090515555606948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,1536,0.009663999908500248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,3072,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,1024,0.008265777594513362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,2560,0.006025777922736273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,512,0.00360000009338061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,768,0.007324444750944774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,512,0.007663999994595845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,512,0.006311999840868845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,2048,0.011062222222487131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,256,0.007350222104125553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,256,0.005587555468082428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,128,0.002899555489420891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,128,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,3072,0.014312888185183207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,64,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,32,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,64,0.0053084443012873335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,32,0.0052791109515561005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,1024,0.008351999852392409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,51200,0.046930667426851064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,65536,0.20377777682410347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,768,0.008036444584528605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,16384,0.01294222225745519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,12288,0.012418666647540199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,51200,0.1604106691148546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,256,0.003539555602603488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,10240,0.012421333127551608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,8192,0.01108088923825158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,12288,0.041217777464124895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,7168,0.010436444646782346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,10240,0.034649779399236046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,6144,0.009688888986905416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,7168,0.025734222597546045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,128,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,6144,0.022900443938043382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,4096,0.007795555724038019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,16384,0.05519022213088142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,5120,0.01998044384850396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,3584,0.007290666302045186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,65536,0.06201955344941881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,3072,0.006719111154476802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,2560,0.00592533333433999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,768,0.003875555677546395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,2048,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,2560,0.012481778032249875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,1536,0.004881777697139316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,5120,0.009110222260157267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,1024,0.004111111164093018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,1536,0.00959822204377916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,768,0.003641777982314428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,1024,0.00797333319981893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,512,0.00332622230052948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,4096,0.017165333032608032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,3584,0.011395555403497485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,768,0.007291555404663086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,8192,0.028487109475665625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,128,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,512,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,64,0.0028488888508743713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,256,0.005614222337802251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,2048,0.011071110765139261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,32,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,128,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,65536,0.05748089154561361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,64,0.005241777747869492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,51200,0.046032889021767505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,16384,0.01275466630856196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,65536,0.20563466019100615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,3584,0.01574133336544037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,51200,0.16047022077772352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,256,0.003192000091075897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,16384,0.05512711074617174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,10240,0.009736888938479954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,8192,0.008896888958083259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,7168,0.008979555633332994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,12288,0.04091644287109375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,10240,0.03462310963206821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,6144,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,8192,0.028584887584050495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,5120,0.01070844464831882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,7168,0.025495110286606684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,3584,0.008427555362383524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,6144,0.022808889547983806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,3072,0.007712889048788283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,4096,0.017183999220530193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,2560,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,3584,0.015504888362354703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,2048,0.005232000102599462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,3072,0.014148443937301636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,1536,0.004557333472702238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,2560,0.012842666771676807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,1024,0.003920000046491623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,2048,0.011242667006121742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,768,0.00388711111413108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,1536,0.009708444277445475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,512,0.003530666646030214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,1024,0.008020444048775567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,256,0.0031822222388452957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,5120,0.01994222237004174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,768,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,128,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,12288,0.01037599974208408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,512,0.006261333409282897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,32,0.002557333351837264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,256,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,128,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,64,0.0052284445199701525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,32,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,4096,0.009016888837019602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,32,0.005165333135260476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,64,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,3072,0.014175110393100314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,16384,1.0245271258884006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,16384,3.2347058190239797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,16384,3.6926951938205295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,51200,3.8749821980794272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,12288,2.552415211995443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,16384,3.0681760576036243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,12288,0.7460266749064127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,12288,2.5453323788113065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,10240,2.0378345913357205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,10240,0.6227991316053603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,10240,2.115623050265842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,8192,1.6293261845906575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,12288,2.3466648525661893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,10240,1.9011234707302516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,8192,1.6282764010959203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,8192,0.5090248849656847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,51200,10.859972635904947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,7168,1.357078234354655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,7168,0.48711289299858945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,7168,1.4090213775634766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,6144,1.4011786778767903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,8192,1.5008541742960613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,7168,1.3192631403605144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,6144,0.4019635518391927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,6144,1.1303200191921658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,5120,0.9965431425306531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,5120,0.9968283971150717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,6144,1.1993466483222115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,4096,0.7724711100260416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,5120,0.9449564615885416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,51200,14.521860758463541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,5120,0.34034045537312824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,4096,0.784729798634847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,3584,0.2561946709950765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,3584,0.6948569085862902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,3072,0.6102364328172472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,4096,0.2851217852698432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,3584,0.6560506820678711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,3072,0.5953928629557291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,3072,0.22444534301757812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,3584,0.6676204469468859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,51200,10.615734524197048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,2560,0.4373404449886746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,3072,0.5520675447252061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,2560,0.1969146728515625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,2560,0.5142951011657715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,2048,0.43319378958808047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,2560,0.47241867913140195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,2048,0.16603733433617487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,2048,0.38284532229105633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,1536,0.34447466002570254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,1024,0.21811289257473418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,1536,0.30475378036499023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,1024,0.1078266633881463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,1024,0.24545245700412324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,1024,0.24704265594482422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,2048,0.4194222291310628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,768,0.09268089135487874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,1536,0.13407022423214382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,4096,0.7230337990654839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,768,0.18731821907891166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,512,0.08397599723603989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,512,0.16893243789672852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,512,0.15140533447265625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,768,0.17984355820549858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,1536,0.28195643424987793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,128,0.07099555598364936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,256,0.07446489069196913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,256,0.12063733736673991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,64,0.0649991101688809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,128,0.1419937743080987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,128,0.09273866812388103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,32,0.06781066788567437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,256,0.09566666682561238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,64,0.14473777347140843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,32,0.14537599351671007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,256,0.15051821867624918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,768,0.1967324415842692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,128,0.07174222336875068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,512,0.1345004505581326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,51200,3.094328986273872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,65536,3.9649874369303384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,16384,3.238287183973524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,51200,7.987676832411025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,16384,2.9526240030924478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,65536,10.311458163791233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,16384,0.7764426867167155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,12288,2.188407050238715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,12288,0.6358799934387207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,16384,2.4097768995496964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,12288,1.9601048363579645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,51200,12.048786587185331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,12288,1.743984858194987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,10240,0.5048729048834907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,51200,8.675433688693577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,10240,1.818695068359375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,10240,1.6499386893378363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,65536,14.802147759331596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,10240,1.5136470794677734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,8192,1.4100888570149739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,8192,1.3252222273084853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,7168,0.3515742090013292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,7168,1.3378453784518773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,8192,1.161410649617513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,7168,1.0891635682847765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,65536,10.941875881618925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,6144,0.3134168783823649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,7168,1.040503078036838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,6144,0.947646247016059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,5120,0.7585875723097059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,5120,0.275947544309828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,8192,0.41770132382710773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,6144,0.9271457460191516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,4096,0.6729351149664985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,5120,0.8540550867716471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,4096,0.22619732220967612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,4096,0.6296639972262913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,3584,0.5021208657158746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,3584,0.20230755541059706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,3584,0.5743298000759548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,4096,0.5842889149983724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,3584,0.5003315607706705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,3072,0.17655821641286215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,3072,0.4748533566792806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,2560,0.43138665623135036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,3072,0.43648444281684023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,5120,0.6928630934821235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,2560,0.15972177187601724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,2048,0.3488231235080295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,2560,0.4007564385732015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,2560,0.36847376823425293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,3072,0.4382186730702718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,1536,0.23439555697970918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,6144,0.8502675692240397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,1024,0.1525768836339315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,1536,0.107951111263699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,1536,0.256677336162991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,768,0.11556711461808945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,1024,0.08521422412660386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,1024,0.19314755333794487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,1024,0.1837573316362169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,768,0.15510222646925184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,768,0.07571199867460462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,2048,0.3326808876461453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,512,0.1026195552614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,2048,0.3027751180860731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,256,0.08557599782943726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,512,0.13322399722205266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,256,0.09583644734488593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,256,0.11897244718339707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,768,0.14739289548661974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,128,0.05559644434187147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,256,0.06034577555126614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,512,0.06780978043874104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,128,0.05838311380810208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,64,0.0465866658422682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,128,0.11230399873521592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,32,0.06272444460127088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,64,0.11423733499315049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,2048,0.1349875529607137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,32,0.11474043793148464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,512,0.1207617786195543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,128,0.07440266344282362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,65536,1.1752693388197157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,1536,0.23951021830240884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,51200,2.750087102254232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,65536,3.4026951260036893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,51200,0.9170764287312826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,16384,0.30494221051534015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,16384,0.8940639495849609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,16384,0.7351973321702746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,12288,0.6919617652893066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,51200,4.044511159261067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,51200,2.627107620239258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,12288,0.6260222329033746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,10240,0.5378915468851725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,16384,0.7802924580044217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,12288,0.5398942099677192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,10240,0.21449955304463705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,10240,0.5465226703219944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,65536,4.778012593587239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,10240,0.45966042412651914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,8192,0.1382639937930637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,12288,0.21020000510745576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,7168,0.38585511843363446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,65536,3.3670152028401694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,8192,0.3658133347829183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,7168,0.12417422400580512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,6144,0.32203910085890025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,6144,0.10854222377141316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,5120,0.27975556585523814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,6144,0.3111911084916857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,6144,0.27811021274990505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,5120,0.09549778037601048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,4096,0.20668089389801025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,5120,0.24179199006822374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,7168,0.3628675672743056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,4096,0.2073955535888672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,4096,0.190665774875217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,7168,0.3183013333214654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,8192,0.3715662161509196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,3584,0.06924711333380805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,3584,0.18182133303748238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,8192,0.4367164505852594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,3584,0.1687164439095391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,5120,0.2600177658928765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,2560,0.1304711103439331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,3072,0.15884799427456325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,3072,0.1501715580622355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,4096,0.07959555255042182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,2048,0.10803377628326416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,2560,0.1264151069853041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,2048,0.1116666661368476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,2048,0.047261334127850003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,1536,0.07933066950903998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,3072,0.06215911441379123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,1536,0.08926044570075141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,1024,0.052870220608181424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,1536,0.039613333013322614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,2560,0.054626666837268405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,1536,0.08653244707319473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,768,0.04380444354481167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,1024,0.06807644499672784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,3072,0.1324631108178033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,768,0.03073599934577942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,512,0.03397866752412584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,2048,0.10673422283596462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,3584,0.18338844511244032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,512,0.04770044485727946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,512,0.046184000041749745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,512,0.02789688772625393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,1024,0.03346844514211019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,1024,0.06729422012964885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,128,0.022340445054901972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,256,0.04263822237650553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,256,0.025432000557581585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,256,0.0369102226363288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,64,0.019018666611777413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,768,0.05559911330540975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,128,0.02457511093881395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,32,0.020103999310069613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,128,0.02998488810327318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,2560,0.1355031066470676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,64,0.04071911176045736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,256,0.02498133314980401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,768,0.05503377649519178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,128,0.04010666741265191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,32,0.040864888164732195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,65536,2.5881288316514754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,65536,0.8907306459214953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,51200,2.1760951148139105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,16384,0.5887644555833604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,51200,0.7000142203436958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,16384,0.6550070974561903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,65536,3.983582390679253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,16384,0.1991786691877577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,65536,2.5040018293592667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,51200,2.7009457482231984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,12288,0.5690044297112359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,51200,1.9309982723659938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,12288,0.4726746877034505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,12288,0.402679098976983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,10240,0.40630398856268984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,10240,0.14089777734544542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,8192,0.3275875515407986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,10240,0.39542754491170246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,8192,0.10577599869834052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,16384,0.5604266590542263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,7168,0.28628354602389866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,8192,0.3160088857014974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,8192,0.2745430999332004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,12288,0.15234043863084581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,6144,0.24426666895548502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,10240,0.41751376787821454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,7168,0.28498933050367564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,6144,0.08254488971498278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,6144,0.2396142217848036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,4096,0.15636266602410212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,6144,0.20885866218143037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,5120,0.07129866547054715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,5120,0.18076354927486846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,4096,0.05973688761393229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,4096,0.15916800498962402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,4096,0.14503910806443956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,3584,0.05396177702479892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,3584,0.12953600618574354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,3072,0.1218515502081977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,5120,0.19701155026753744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,3072,0.12234221564398871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,3584,0.1453733311759101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,3072,0.0487040016386244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,7168,0.09818400277031793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,2560,0.04371910956170824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,7168,0.24172178904215494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,2048,0.07950577470991346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,2560,0.09865688615375096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,2048,0.03755022088686625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,2048,0.08681422472000122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,1536,0.06259733438491821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,3072,0.11528089311387803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,1536,0.031256887647840716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,5120,0.20159644550747344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,1024,0.04170133339034187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,1024,0.05305511090490553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,1024,0.05346044566896227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,768,0.0323884454038408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,1024,0.025016888976097107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,2560,0.0926026635699802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,768,0.02304977840847439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,768,0.04209422071774801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,512,0.025042666329277888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,3584,0.1395662228266398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,1536,0.06715999709235297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,256,0.0255386663807763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,512,0.021016000045670405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,512,0.03478399912516276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,512,0.03686666819784377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,128,0.02232355541653103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,256,0.030947556098302204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,256,0.029436445898479883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,128,0.029551111989551123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,64,0.016460445192125108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,2560,0.1049191090795729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,32,0.016640888320075143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,64,0.029495110114415485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,32,0.029176887538697984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,2048,0.08333777719073826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,128,0.019112888309690688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,128,0.024307555622524683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,768,0.0450853341155582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,256,0.018920888503392536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,65536,0.7453449037339953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,1536,0.0688248872756958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,65536,2.3278107113308377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,51200,1.7247822019788954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,16384,0.5365164544847276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,51200,0.5902738041347927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,16384,0.17019377814398873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,65536,2.9997520446777344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,12288,0.440391116672092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,16384,0.46976179546780056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,65536,2.19505861070421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,12288,0.4151742193433974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,12288,0.13435199525621203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,51200,1.624222225613064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,10240,0.11971999539269342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,16384,0.566497802734375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,10240,0.29309333695305717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,8192,0.27727556228637695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,8192,0.2659440040588379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,8192,0.09272088607152303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,12288,0.3501075638665094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,8192,0.23746310340033638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,6144,0.20181422763400605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,10240,0.3325795597500271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,7168,0.2100906636979845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,6144,0.2006657785839505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,6144,0.17805066373613146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,5120,0.1588008933597141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,10240,0.3475671079423692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,7168,0.22636622852749297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,4096,0.13028621673583984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,5120,0.1680062214533488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,51200,2.2869750128851996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,5120,0.14983733495076498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,4096,0.051390220721562706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,3584,0.11405332883199056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,4096,0.1352133353551229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,4096,0.12439110543992783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,6144,0.0731368859608968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,3584,0.04600888821813795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,3584,0.11847110589345296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,3584,0.11073955562379624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,2560,0.08407555686102973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,7168,0.08320355415344238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,3072,0.09911111328336929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,2560,0.03729155659675598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,2560,0.08968266513612534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,2048,0.06598399745093451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,2560,0.08505422539181179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,3072,0.100509332285987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,2048,0.03164533442921109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,1536,0.055812444951799184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,2048,0.07243111398484972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,3072,0.1043928861618042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,1024,0.03146400054295858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,1536,0.025787555509143408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,1536,0.05848177936342028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,1024,0.02180800007449256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,768,0.028579556279712256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,1024,0.04610933197869194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,2048,0.0740284456147088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,768,0.036173333724339805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,512,0.02425333360830943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,1536,0.0587715572781033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,7168,0.23379378848605684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,512,0.02978844443957011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,256,0.01906222270594703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,512,0.032551109790802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,256,0.026519111461109583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,256,0.02568000058333079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,768,0.03882133298450046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,512,0.01828533411026001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,1024,0.04495911134613884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,128,0.01664444473054674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,64,0.015790222419632804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,5120,0.06339289082421197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,64,0.025274667474958632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,128,0.02516266703605652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,32,0.02479466630352868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,128,0.02167822255028619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,256,0.01701955497264862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,32,0.015768888923856948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,65536,0.6329893536037868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,65536,1.9800906711154516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,3072,0.04154577851295471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,51200,1.4026124742296007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,65536,2.41595458984375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,16384,0.4206844435797797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,65536,1.6993484497070312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,768,0.01943555474281311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,16384,0.15421422322591147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,12288,0.2801617781321208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,16384,0.3933528794182672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,51200,1.9612168206108942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,51200,1.2996898227267795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,12288,0.11374399397108291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,51200,0.4875786569383409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,10240,0.2904471026526557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,12288,0.28139376640319824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,128,0.01663377715481652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,10240,0.27241600884331596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,10240,0.24397778511047363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,8192,0.0759484436776903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,8192,0.21663999557495117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,8192,0.19456888569725883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,12288,0.3316640059153239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,7168,0.06940711206860013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,7168,0.19204978148142496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,7168,0.169669336742825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,10240,0.09402755233976577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,8192,0.231170654296875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,6144,0.05955288807551066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,6144,0.1482693354288737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,7168,0.20650133821699354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,5120,0.13719377252790663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,5120,0.0528728895717197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,4096,0.10866666502422756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,6144,0.17756711112128365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,16384,0.4665573438008626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,4096,0.043143110142813794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,5120,0.14378844367133245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,3584,0.09772355688942803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,3584,0.0938791102833218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,3072,0.07872177494896783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,3584,0.03823111123508877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,3072,0.034267554680506386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,3072,0.08612977796130711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,3072,0.08403911193211873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,2560,0.07396622498830159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,2560,0.07267733414967854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,2560,0.03070666723781162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,4096,0.11000355747010972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,2560,0.06791200240453084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,3584,0.10017066531711155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,6144,0.16355378097958034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,1536,0.045647998650868736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,2048,0.026491555902693007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,2048,0.06127999888526069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,1536,0.021943999661339655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,1536,0.05018844538264804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,5120,0.12699822584788004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,1024,0.03759288787841797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,768,0.023043556345833674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,2048,0.061490668190850147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,4096,0.10717866818110149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,768,0.016745777593718637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,768,0.03040800160831875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,512,0.01923022170861562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,768,0.03331999977429708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,2048,0.05541244480344984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,512,0.016121778223249648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,512,0.025172443853484258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,512,0.02827377782927619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,256,0.022461333208613928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,128,0.011752000285519494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,256,0.022099556194411382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,128,0.021312000022994146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,128,0.018937778141763475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,64,0.010027555955780877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,256,0.014133334159851074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,32,0.011942221886581846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,1024,0.018248889181349013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,64,0.021254221598307293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,1024,0.03910933269394769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,32,0.021035555336210463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,1024,0.03159022331237793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,128,0.014589332871966891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,65536,0.5341253280639648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,256,0.014862222803963555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,51200,1.2314942677815754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,51200,0.4192186726464166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,65536,1.9973778194851344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,16384,0.4126613405015733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,65536,1.5207031038072374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,65536,1.5179804695977106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,51200,1.7306515375773113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,16384,0.1470186710357666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,16384,0.41274311807420516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,51200,1.1639155281914604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,12288,0.09948711262808906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,1536,0.04888888862397936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,12288,0.28859466976589626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,12288,0.26759645673963756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,10240,0.08497689167658488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,8192,0.17029421859317354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,10240,0.24966221385531953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,10240,0.2179351117875841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,8192,0.06737422280841403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,8192,0.19340711169772676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,12288,0.28038401073879665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,16384,0.3417333232031928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,6144,0.14509688483344182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,10240,0.22440889146592882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,7168,0.1540586683485243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,6144,0.0528657767507765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,6144,0.1463599999745687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,6144,0.13550133175320095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,7168,0.15780888663397893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,5120,0.12200444274478489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,4096,0.08665066957473755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,5120,0.11439555221133763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,7168,0.06081866555743747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,7168,0.16842399703131783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,4096,0.03817777832349142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,4096,0.09659910864300197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,5120,0.10541777478324042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,3584,0.0875120030509101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,3584,0.03478755553563436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,3584,0.08522488673528035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,5120,0.046338667472203575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,3072,0.07701066467497084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,8192,0.17419200473361543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,3072,0.030924446052975122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,3584,0.08625244432025486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,2560,0.06634222136603461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,2560,0.028245333168241713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,2048,0.044687112172444664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,2560,0.06529066960016887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,3072,0.06620799832873873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,1536,0.037111110157436795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,2048,0.024158222807778254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,2048,0.055367112159729004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,2048,0.05525066786342197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,1024,0.025460445218616064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,1536,0.044065776798460216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,1536,0.04560710986455282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,4096,0.09863555431365967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,768,0.020253333780500624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,1024,0.03362933463520474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,1024,0.035869333479139544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,512,0.015251555376582675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,768,0.015850666496488783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,768,0.027828445037206013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,512,0.02276177704334259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,256,0.012077333198653327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,512,0.025859556264347498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,256,0.019949333535300363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,1024,0.018183110488785636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,256,0.020283554991086323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,256,0.013785777820481194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,768,0.030611554781595867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,128,0.019267555740144517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,64,0.009727111293209923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,512,0.014582221706708273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,32,0.00997866690158844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,64,0.018961777289708454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,32,0.018910222583346896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,128,0.011114666859308878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,128,0.017774222625626456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,1536,0.020794666475719877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,65536,1.2318141725328233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,65536,0.49491553836398655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,65536,1.7358924018012152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,3072,0.07635200023651123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,65536,1.3313324186537001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,16384,0.33262043529086643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,51200,0.3862008783552382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,51200,1.3185759650336373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,16384,0.35747644636366105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,16384,0.11926310592227513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,12288,0.24675467279222277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,51200,1.1165182325575087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,16384,0.30481333202785915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,2560,0.05659733215967814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,12288,0.08847822083367242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,12288,0.2523724502987332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,12288,0.22667821248372397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,10240,0.07582577731874254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,8192,0.18758311536577013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,10240,0.21116977267795137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,10240,0.19127733177608916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,7168,0.1517582204606798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,8192,0.1660346719953749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,8192,0.15425066153208414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,51200,0.9850373797946506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,128,0.013839999834696451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,7168,0.05489244394832187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,7168,0.1370799938837687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,5120,0.10787733395894368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,6144,0.11948622597588433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,8192,0.06536977820926242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,5120,0.10726933346854316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,5120,0.040960888067881264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,5120,0.10203555557462905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,6144,0.13026044103834364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,10240,0.22248355547587076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,4096,0.08683288759655422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,3584,0.07487466600206164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,4096,0.03426400158140395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,4096,0.08467644453048706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,3072,0.06460355387793647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,3584,0.03086044390996297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,3584,0.07704710960388184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,3584,0.0762080020374722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,3072,0.028296000427669946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,4096,0.08086044258541532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,3072,0.06755022207895915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,3072,0.06794755326377021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,2048,0.04413511024581062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,2560,0.025320000118679468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,2560,0.05823555257585314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,2560,0.05917244487338596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,6144,0.12811199824015299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,1536,0.03478044602606032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,2048,0.04864977796872457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,2048,0.04973422156439888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,1536,0.018445332845052082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,1536,0.04145777887768216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,2560,0.053764445914162524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,1024,0.03014933400683933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,7168,0.14740978346930608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,1024,0.01550222271018558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,768,0.02132533325089349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,2048,0.021892444954978094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,768,0.01479555500878228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,1024,0.025184889634450276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,512,0.02050933407412635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,1024,0.03291555576854282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,6144,0.047932446002960205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,256,0.014387556248241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,512,0.023036445180575054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,768,0.024691555235120986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,512,0.018019555343521964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,256,0.017733333839310538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,256,0.012799999780125089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,256,0.018597332967652213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,128,0.012850667039553324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,64,0.011596444580290051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,128,0.017092444830470614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,32,0.011901333100265928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,512,0.013615999784734515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,64,0.017291555802027386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,32,0.016929777132140267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,1536,0.03894133369127909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,128,0.012474666866991254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,128,0.01721333298418257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,65536,0.43401156531439883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,65536,1.0118693245781791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,51200,0.8756808704800076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,16384,0.2798302173614502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,768,0.027816000911924574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,65536,1.5119689305623372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,51200,1.1617839601304796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,65536,1.144985728793674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,16384,0.39922576480441624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,16384,0.2678595648871528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,51200,0.887497795952691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,12288,0.09926133023367988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,12288,0.22030933698018393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,10240,0.1822213331858317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,12288,0.20418222745259604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,10240,0.06676177846060859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,10240,0.1937111086315579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,10240,0.16729244920942518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,8192,0.14303821987575954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,16384,0.10406311353047688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,8192,0.1366248925526937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,7168,0.04779288834995694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,7168,0.12628532780541316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,6144,0.11021066374248928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,7168,0.12056533495585124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,8192,0.13503911760118273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,5120,0.08489333258734809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,51200,0.3522871070437961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,6144,0.10957333114412095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,6144,0.10511555274327596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,7168,0.11965600649515788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,5120,0.03548177745607164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,12288,0.20439110861884224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,5120,0.09086400270462036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,4096,0.030245333909988403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,3584,0.05829422341452705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,4096,0.07510311073727079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,4096,0.07544710901048449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,3584,0.027569777435726587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,3072,0.05174755718972948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,3584,0.06657155354817708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,6144,0.041466666592492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,3584,0.06779111093944974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,2560,0.044870220952563815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,3072,0.024895111719767254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,3072,0.05868177943759494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,8192,0.0532533327738444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,3072,0.05952088700400459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,2560,0.022543110781245764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,2560,0.05045333173539904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,1536,0.027811557054519653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,2048,0.01984355515903897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,2048,0.04313244422276815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,2048,0.044450667169359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,1024,0.019924443629052904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,1536,0.036907556984159685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,1024,0.026338666677474976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,4096,0.07111022207472059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,2048,0.033336000310050115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,5120,0.09217333131366306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,1024,0.014871110518773397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,2560,0.05186133252249824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,512,0.012296000288592445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,768,0.013864888913101621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,768,0.02159022291501363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,768,0.02491644521554311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,512,0.012711111042234631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,512,0.017934223016103108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,512,0.02084000077512529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,256,0.015722667177518208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,256,0.012033777932325998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,256,0.01726222203837501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,1024,0.02938488788074917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,128,0.015199111567603217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,128,0.015217777755525378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,64,0.009355555805895064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,128,0.012142221960756512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,32,0.009829333258999718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,64,0.01514844430817498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,128,0.009002666506502364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,32,0.014861333701345654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,1536,0.017337777548366122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,256,0.010091555615266165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,65536,1.0894222259521484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,65536,0.356931553946601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,65536,1.2196541892157662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,51200,0.7153244548373752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,16384,0.2588053279452854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,65536,0.9449218114217123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,1536,0.03436355458365546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,51200,0.9251306321885852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,16384,0.24438844786749947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,12288,0.1919306649102105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,16384,0.08629777696397568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,16384,0.23452266057332358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,12288,0.06505955590142144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,768,0.015791111522250705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,10240,0.1613022221459283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,12288,0.17002667321099174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,10240,0.05654755565855238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,10240,0.1494551102320353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,51200,0.28346488210890025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,8192,0.04482044445143806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,51200,0.7155991130405002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,8192,0.11735022068023682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,7168,0.0399271117316352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,7168,0.10680355628331502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,6144,0.09371111128065322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,7168,0.10467111402087742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,8192,0.13430844412909612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,6144,0.03525422347916497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,6144,0.09229422277874416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,5120,0.08068266842100355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,6144,0.09100177552964951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,8192,0.1191271146138509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,4096,0.06137333313624064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,5120,0.030921777089436848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,5120,0.07738488912582397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,5120,0.07852266894446479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,3584,0.05095200075043572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,4096,0.026343110534879897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,4096,0.06319200330310397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,4096,0.06510400109820895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,12288,0.18887110551198324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,3584,0.056461334228515625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,3584,0.059169775909847684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,2560,0.03692977627118429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,3072,0.050118221176995166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,3072,0.05254044466548496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,7168,0.10717244280709161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,2048,0.030441777573691473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,2560,0.019998222589492798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,2560,0.045316444502936475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,2048,0.0369466675652398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,3584,0.02390044430891673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,1536,0.023582221733199224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,3072,0.022030222747060988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,10240,0.14379999372694227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,1536,0.029440889755884807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,1024,0.01809155609872606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,1536,0.015105777316623263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,2560,0.043311112456851535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,1024,0.01349688900841607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,1024,0.022276444567574397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,1024,0.025379555092917547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,2048,0.01723377737734053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,2048,0.038784000608656145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,768,0.012772444221708508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,768,0.021951999929216173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,512,0.01566133399804433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,1536,0.03327288892534044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,768,0.014642667439248828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,256,0.013591110706329346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,3072,0.042945779032177396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,256,0.011066666907734342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,256,0.015153777268197803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,512,0.01179377817445331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,512,0.018752889500723947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,128,0.013895110951529609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,128,0.011053333679835001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,32,0.0088755554623074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,64,0.013117333253224691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,512,0.011686222420798408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,32,0.013241777817408243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,128,0.013144888811641268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,256,0.009047110875447592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,64,0.008311111066076491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,65536,0.7462115817599826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,65536,0.30131644672817653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,51200,0.5724595387776693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,768,0.01867822143766615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,16384,0.18971555762820771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,65536,1.057071155971951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,51200,0.7836186620924207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,51200,0.6089279916551378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,16384,0.20008799764845106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,12288,0.1414711078008016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,128,0.007341333561473423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,12288,0.16325155893961588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,51200,0.23983822928534615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,10240,0.12386577659183079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,12288,0.05721600188149346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,12288,0.14438045024871826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,8192,0.09522933430141872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,10240,0.13377422756618923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,10240,0.12325333224402533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,8192,0.10799022515614827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,7168,0.08316889074113634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,7168,0.09572622511121963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,10240,0.049689776367611356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,65536,0.8090782165527344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,7168,0.0355751117070516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,7168,0.08988622162077163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,8192,0.10421688689125909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,6144,0.08311733272340563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,6144,0.03142933381928338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,5120,0.05872088670730591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,5120,0.0278479986720615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,5120,0.07040977478027344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,5120,0.06764977508121066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,16384,0.21340889400906035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,6144,0.07379466957516141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,4096,0.05785955323113335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,3584,0.043064888980653554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,8192,0.0391244457827674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,3584,0.021535999245113794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,6144,0.07949066824383207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,3072,0.03641244437959459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,3584,0.05165511038568285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,4096,0.048375109831492104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,3584,0.05175822310977512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,3072,0.019722667005327012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,3072,0.045759999089770846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,4096,0.02375288969940609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,2560,0.03969066672854953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,2560,0.039770666095945574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,2048,0.033448888195885554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,2048,0.03452355662981669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,3072,0.04610044426388211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,2048,0.01589866644806332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,2560,0.018213333355055917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,1536,0.027032888597912256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,4096,0.05721955829196506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,1024,0.014574221438831754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,1536,0.01423911088042789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,1024,0.012440000143316058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,1024,0.020636444290479023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,768,0.012252444194422828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,1024,0.022685334086418152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,2560,0.029512001408471003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,16384,0.07378133138020833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,768,0.01959022217326694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,2048,0.02461599972512987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,1536,0.028860443168216284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,256,0.007351111206743453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,512,0.01685422162214915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,256,0.012799111505349478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,256,0.014360000689824423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,1536,0.01953866746690538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,128,0.006904888898134232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,128,0.012111999922328524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,512,0.014319999350441827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,64,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,128,0.010398222340477837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,512,0.010957332948843638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,32,0.006613333192136552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,64,0.012135999898115793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,32,0.012097777591811286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,768,0.011621333658695221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,256,0.010445333189434476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,65536,0.5753777821858724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,65536,0.2850382328033447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,51200,0.5107066896226671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,65536,0.8681706852383084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,51200,0.23130843374464247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,51200,0.6991173426310221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,16384,0.16948533058166504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,512,0.009690666364298927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,12288,0.12602310710483128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,51200,0.5831795798407661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,16384,0.18933510780334473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,16384,0.188719113667806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,128,0.01348444405529234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,12288,0.05496711201137967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,12288,0.14045688841078016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,12288,0.14642311467064753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,10240,0.11898399723900689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,8192,0.08905244535870022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,10240,0.1210977766248915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,16384,0.07203021976682875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,7168,0.07609155442979601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,10240,0.11042755842208862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,7168,0.08530577686097886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,10240,0.04840088884035746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,65536,0.7617413202921549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,7168,0.087826669216156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,768,0.01737600068251292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,8192,0.0958462225066291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,8192,0.10037955310609604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,6144,0.030135999123255413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,6144,0.07755555709203084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,5120,0.062282668219672314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,5120,0.06673689021004571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,6144,0.06625333097245958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,4096,0.043736000855763756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,8192,0.03711822297837999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,6144,0.07389422257741292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,4096,0.022592888938056097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,3584,0.04547022117508782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,5120,0.026532444689008925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,3072,0.03420977791150411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,3584,0.02055644492308299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,4096,0.05071999960475498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,4096,0.05607910950978597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,3072,0.04059022333886888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,3072,0.04456266760826111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,2560,0.03505333264668783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,7168,0.034625778595606484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,2560,0.017151999804708693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,3584,0.05014311273892721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,3584,0.038648890124426946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,2560,0.028919110695521038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,1536,0.019858666592174105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,2048,0.01553244392077128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,1536,0.02358666724628872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,2560,0.0392524434460534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,1536,0.027279999521043565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,1024,0.015030221806632148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,1536,0.014018666413095264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,1024,0.012135999898115793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,1024,0.018424888451894123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,5120,0.05234488844871521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,1024,0.02204711072974735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,3072,0.018795554836591084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,512,0.00959022260374493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,768,0.011406222151385413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,768,0.015861334072219003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,768,0.01905955539809333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,256,0.007923555870850882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,512,0.010680889089902243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,512,0.016521778371598985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,256,0.011749332977665795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,128,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,768,0.01202755586968528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,2048,0.03371289041307237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,128,0.011168888873524137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,64,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,128,0.01275466630856196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,32,0.006611555400821898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,512,0.0131217779384719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,2048,0.025204444924990337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,64,0.011187555889288584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,32,0.011052444577217102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,256,0.010088889135254754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,256,0.013985777894655863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,65536,0.537543084886339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,65536,0.2639137903849284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,51200,0.44020711051093203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,128,0.01089777797460556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,65536,0.7268302175733777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,51200,0.2106542189915975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,51200,0.5889591111077203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,51200,0.4837084346347385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,16384,0.06718844175338745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,16384,0.16460800170898438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,16384,0.15855377250247532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,2048,0.029495110114415485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,12288,0.051150222619374595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,12288,0.12918666998545328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,12288,0.11730133162604438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,8192,0.07502844598558214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,10240,0.10434399710761176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,16384,0.1408240000406901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,65536,0.6115768750508627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,7168,0.062146663665771484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,10240,0.09983377986484104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,8192,0.08422578043407863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,12288,0.10780621899498834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,8192,0.08283733659320407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,6144,0.0510480006535848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,7168,0.032080888748168945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,10240,0.09310577975379096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,7168,0.07336266835530598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,5120,0.046111110183927745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,6144,0.06474400228924222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,6144,0.06430933210584852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,10240,0.04253066579500834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,4096,0.036686221758524575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,5120,0.024547555380397375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,5120,0.05481333202785916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,3584,0.032920890384250216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,4096,0.021293333835071985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,4096,0.04497777753406101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,4096,0.04696711235576206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,3584,0.040362666050593056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,3584,0.018557333283954196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,7168,0.07438133160273235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,3584,0.04244533181190491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,6144,0.028635554843478735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,2560,0.022808889547983806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,3072,0.035989334185918175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,3072,0.0374257763226827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,5120,0.057166225380367704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,2560,0.031122667921913996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,2048,0.020216888851589628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,2560,0.03301599952909682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,2048,0.014966222974989148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,3072,0.027688889039887324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,2048,0.028554668029149372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,3072,0.017867555220921833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,1536,0.021398221453030903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,1024,0.011893333660231696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,1536,0.012423999607563019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,2560,0.016125332978036668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,1024,0.016516443755891588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,768,0.009670221971140968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,2048,0.02616888946957058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,1536,0.015629332926538255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,512,0.00792622235086229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,768,0.014163555370436775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,768,0.01686311099264357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,1536,0.02375999920898014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,256,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,8192,0.03527822097142538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,1024,0.011062222222487131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,1024,0.018983110785484314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,128,0.005759999983840519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,256,0.013131555583741931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,768,0.010391111175219217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,128,0.010293333066834344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,64,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,512,0.012120889292822944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,32,0.005306666509972678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,512,0.014527110589875115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,64,0.010070222119490305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,32,0.010035555395815108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,256,0.009390222529570261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,65536,0.4818302260504828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,65536,0.21719555060068765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,51200,0.39171555307176376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,65536,0.5804968939887153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,512,0.009936889012654623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,65536,0.5890480147467719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,51200,0.17171733909183076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,256,0.010616000327799054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,51200,0.45307646857367617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,16384,0.14332977930704752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,16384,0.05743288993835449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,16384,0.15215822060902914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,128,0.0120666664507654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,12288,0.044488890303505786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,12288,0.1080568896399604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,10240,0.08229955699708727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,12288,0.1181466711892022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,16384,0.12530488438076443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,10240,0.039064887497160174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,10240,0.0897759993871053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,10240,0.09863822327719794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,7168,0.051057779126697116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,12288,0.09795910782284206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,51200,0.49504889382256406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,7168,0.027615111735132005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,128,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,6144,0.0457822216881646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,7168,0.07242400116390653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,8192,0.05758221944173177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,6144,0.024803555674023096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,5120,0.03713688916630215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,6144,0.06319999694824219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,8192,0.030407110850016277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,5120,0.047372443808449634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,8192,0.08184622393714057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,5120,0.056016001436445445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,4096,0.01887022289964888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,4096,0.039142221212387085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,3584,0.026986666851573523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,4096,0.045965333779652916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,6144,0.05614044268925985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,3584,0.03529599971241421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,3584,0.0421377784676022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,8192,0.07265510823991564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,4096,0.03031911121474372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,2560,0.019900444481107924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,3072,0.037567999627855085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,7168,0.0646408862537808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,2560,0.026871111657884385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,2048,0.016579555140601266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,2560,0.032822221517562866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,3072,0.023860444625218708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,2048,0.022838221655951604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,5120,0.02165688905451033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,1536,0.013648000028398303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,2048,0.027753776974148218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,3072,0.031361778577168785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,3072,0.016000888413853116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,1024,0.010027555955780877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,1536,0.011909333368142446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,1536,0.01885688967174954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,1536,0.023018666439586218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,768,0.008822222550710043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,1024,0.010790222220950656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,1024,0.014661333627170987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,3584,0.01738933391041226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,512,0.006931555353932911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,768,0.010290666586822933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,768,0.012545777691735161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,768,0.016555555992656283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,256,0.00591466658645206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,512,0.009711999860074785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,512,0.014137778017255994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,256,0.009404444032245213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,256,0.012643555800120035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,2560,0.014759999182489185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,1024,0.018668444620238412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,128,0.00903733323017756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,128,0.00905866672595342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,128,0.012044444680213928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,2048,0.013456000222100152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,32,0.005766222046481238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,512,0.010713777608341642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,64,0.009016888837019602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,65536,0.32973066965738934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,256,0.009363555245929295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,128,0.005674666828579373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,65536,0.17414400312635633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,51200,0.27188889185587567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,16384,0.08912977907392715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,64,0.005584888988071018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,51200,0.36950845188564724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,16384,0.11915466520521377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,51200,0.34939556651645237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,12288,0.06867911418279012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,16384,0.11710045072767471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,12288,0.03636711173587375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,12288,0.09068710936440362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,65536,0.4749937587314182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,12288,0.09006577730178833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,65536,0.4429022206200494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,10240,0.03176800078815884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,10240,0.07558488845825195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,10240,0.07674666908052233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,16384,0.04647555616166857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,32,0.009022222624884712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,7168,0.03861422340075175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,8192,0.06400889158248901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,7168,0.022675555613305833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,10240,0.055289778444502086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,6144,0.030825777186287776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,7168,0.056567112604777016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,51200,0.13566399945153132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,6144,0.020263999700546265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,8192,0.0431253347131941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,6144,0.04751999841796028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,6144,0.050083554453319974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,4096,0.02237244447072347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,5120,0.04021866785155402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,5120,0.043663998444875084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,4096,0.03361777795685662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,4096,0.01575199928548601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,3584,0.020023999942673575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,4096,0.036884443627463445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,3584,0.014701333310869006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,3584,0.030527111556794908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,5120,0.027111111415757075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,3584,0.03314666615592109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,5120,0.01806311143769158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,2560,0.015232889188660515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,3072,0.02958577871322632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,7168,0.05424000157250298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,2560,0.023185777995321486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,2048,0.01274400038851632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,2560,0.01259200026591619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,3072,0.017893332574102614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,2048,0.019816888703240287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,1536,0.010758221977286868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,2048,0.011439999772442712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,3072,0.026886221435334947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,1536,0.016546666622161865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,1024,0.008992888861232335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,1536,0.018970666660202872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,2560,0.025975111458036635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,1024,0.012810666528013019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,1024,0.01591466698381636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,2048,0.02217777735657162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,8192,0.06103822257783678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,1536,0.010715555813577441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,512,0.006656000183688269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,512,0.009672889278994666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,768,0.007965332931942409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,512,0.009008888569143083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,256,0.005607999861240387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,8192,0.024977778395016987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,768,0.009378666679064432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,768,0.014157333307796054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,256,0.008352888955010308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,128,0.004912000149488449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,256,0.008524444368150499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,1024,0.009724444813198512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,128,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,3072,0.013807111316257052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,64,0.004574222283230888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,128,0.01039644413524204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,32,0.004564444637960858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,64,0.008016889293988546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,32,0.008029333419269985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,256,0.011394666300879585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,65536,0.16540266407860651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,65536,0.3801040119594998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,51200,0.20524799823760986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,128,0.008435555630260045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,512,0.012506666282812754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,16384,0.06049066781997681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,65536,0.20144355297088623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,51200,0.2977742354075114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,16384,0.04441510968738132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,12288,0.04758400056097242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,16384,0.11651021904415554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,12288,0.03398488958676656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,768,0.01111644423670239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,51200,0.13042488363054064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,10240,0.061547557512919106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,51200,0.34804532263014054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,8192,0.03105866577890184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,10240,0.07609955469767253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,12288,0.07270577881071302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,65536,0.4377395576900906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,8192,0.06372621986601087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,8192,0.023248889380031165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,7168,0.04448355568779839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,16384,0.09674044450124104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,7168,0.05590044127570259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,10240,0.030633777379989624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,7168,0.021344888541433547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,6144,0.019280888968043856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,6144,0.03899911046028137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,6144,0.048920889695485435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,12288,0.08904266357421875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,7168,0.026586666703224182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,4096,0.017210667332013447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,5120,0.01699733402993944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,5120,0.042555554045571216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,4096,0.02750844425625271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,4096,0.015215999550289579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,4096,0.03571555680698819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,8192,0.04981244603792826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,3584,0.02522666586769952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,10240,0.03973333372010125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,3584,0.03184444374508328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,3584,0.014166222678290473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,5120,0.03330311179161072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,3072,0.013119999733236102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,2560,0.011815110842386881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,3072,0.022310222188631695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,3072,0.02824266751607259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,2048,0.010324444207880232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,2560,0.012082666986518435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,3584,0.015263999501864115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,2560,0.02496266696188185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,1536,0.008432000047630733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,2048,0.016872889465755887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,5120,0.020605333977275424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,2048,0.02180622186925676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,1024,0.007079111205206976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,1536,0.010393777655230628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,1536,0.01847822301917606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,1024,0.011226666470368704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,768,0.005991999887757831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,6144,0.023386667172114056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,2560,0.01984355515903897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,2048,0.0110871113008923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,512,0.005486222190989389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,768,0.013870221873124441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,3072,0.01369066701995002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,1536,0.014143110977278816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,256,0.004562666846646203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,512,0.008672888908121321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,512,0.012154666913880242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,1024,0.009393778112199571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,1024,0.01516533394654592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,256,0.008376888930797577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,256,0.0107360002067354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,768,0.009752000371615091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,128,0.0069857777820693115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,64,0.004562666846646203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,128,0.010427555276287926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,64,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,32,0.007019555403126611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,65536,0.19758933120303682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,128,0.003929777691761653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,768,0.009111999637550777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,65536,0.29172711902194554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,512,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,51200,0.2568968931833903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,256,0.0073626670572492815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,51200,0.22835644086201987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,51200,0.12860000133514404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,128,0.008367111285527548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,16384,0.04963733421431648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,32,0.00489688871635331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,16384,0.04384711053636339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,16384,0.08585244417190552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,16384,0.07946133613586426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,12288,0.0647182199690077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,10240,0.03017066584693061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,65536,0.16142400105794272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,65536,0.3267742262946235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,10240,0.054535110791524254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,51200,0.15337955951690674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,10240,0.03021777669588725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,12288,0.036631110641691424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,10240,0.05266044537226359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,8192,0.02311555544535319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,8192,0.04447022080421448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,8192,0.0444471107588874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,7168,0.021083555287784998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,6144,0.01940355532699161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,7168,0.039481778939565025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,6144,0.018969777557584975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,5120,0.017239999439981248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,8192,0.02409066590997908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,7168,0.022276444567574397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,5120,0.029696000946892634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,5120,0.01699733402993944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,12288,0.03340977761480544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,7168,0.03963022099600898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,4096,0.014875556031862894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,3584,0.012896888785892062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,6144,0.03501155641343858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,3584,0.022462222311231825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,3584,0.01385155568520228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,3072,0.011462222370836469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,12288,0.06154844495985243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,4096,0.024870221813519795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,3072,0.020230222079488967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,3072,0.021186666356192693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,3072,0.01312711089849472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,2560,0.017889777819315594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,2560,0.018922666708628338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,2048,0.008722667064931657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,2048,0.015465777781274585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,4096,0.026398221651713055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,2560,0.00998044427898195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,1536,0.007299555672539606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,2048,0.011117333339320289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,2048,0.016703999704784818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,1536,0.01277511070171992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,3584,0.023759111762046814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,1536,0.014716444744004143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,1024,0.010269333091047075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,1024,0.012467555701732635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,768,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,768,0.009105777574910058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,4096,0.014246222045686511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,2560,0.0122097780307134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,512,0.004948444250557157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,768,0.00905955582857132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,768,0.011413333316644033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,6144,0.034834666384591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,512,0.007810666329330868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,1024,0.009402666654851701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,512,0.01038044442733129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,256,0.008362666600280339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,256,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,128,0.003938666648334927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,256,0.009380444056457942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,128,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,64,0.003875555677546395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,128,0.00889866633547677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,32,0.003879111260175705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,512,0.00867377801073922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,256,0.004219555606444676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,64,0.006613333192136552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,65536,0.15046221680111355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,32,0.006609777609507243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,51200,0.11560800340440537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,65536,0.11824799908532037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,5120,0.03024622135692173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,65536,0.29016799396938747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,16384,0.03987911012437608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,51200,0.09466666645473903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,51200,0.22459912300109863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,51200,0.22521689203050402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,16384,0.07543200254440308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,16384,0.03305600086847941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,16384,0.07817688915464613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,12288,0.025906667113304138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,1536,0.010393777655230628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,65536,0.2852062119377984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,10240,0.047704888714684375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,12288,0.02787999974356757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,10240,0.023595554961098567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,8192,0.018361777067184448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,12288,0.05719733238220215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,1024,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,128,0.006613333192136552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,7168,0.01663466625743442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,8192,0.017286222842004564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,10240,0.05229155553711785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,7168,0.03529333406024509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,6144,0.014851555228233337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,7168,0.015896000795894198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,8192,0.039389332135518394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,10240,0.022833777798546687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,6144,0.03087644444571601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,5120,0.013467555244763693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,6144,0.014559111661381192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,6144,0.03465155429310269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,4096,0.011052444577217102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,5120,0.013282666603724161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,5120,0.026491555902693007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,5120,0.02948533495267232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,3584,0.010416888528399997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,4096,0.022281777527597215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,7168,0.038768887519836426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,4096,0.0255377772781584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,3072,0.009176889227496253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,3584,0.020228443874253165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,12288,0.06051555607053968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,3072,0.01811022228664822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,2560,0.008621333373917473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,3072,0.020389333367347717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,2560,0.010103999740547603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,2560,0.015847999188635085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,2560,0.018534221582942538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,3584,0.011408000356621213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,2048,0.013831111292044321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,3584,0.02327022287580702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,1536,0.006641777853171031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,2048,0.009687999884287516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,2048,0.016359110673268635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,3072,0.010741333166758219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,1024,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,1536,0.009036444127559662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,1536,0.011668444507651858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,768,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,1024,0.008681777450773451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,1024,0.009422221945391761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,2048,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,512,0.004214222232500712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,768,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,768,0.011072888970375061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,8192,0.04354222284422981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,512,0.007330666813585494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,512,0.008001777860853408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,512,0.010086221827401055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,256,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,256,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,128,0.003560888684458203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,256,0.009011555876996782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,4096,0.011767110890812345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,1024,0.012427555190192329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,128,0.007360888852013483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,128,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,768,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,128,0.00867022242810991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,65536,0.08043911059697469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,64,0.005981333139869902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,256,0.0037297776175869834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,51200,0.06665155622694227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,65536,0.11674577660030788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,65536,0.24222132894727919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,1536,0.014230221509933472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,16384,0.02605244517326355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,51200,0.09416533178753322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,51200,0.19120000468360054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,51200,0.2149991061952379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,32,0.003546666767862108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,16384,0.03294577863481309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,16384,0.0633680025736491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,16384,0.07458666960398357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,32,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,10240,0.02052533295419481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,12288,0.04889600144492256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,12288,0.05782133340835571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,8192,0.01589866644806332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,10240,0.022219555245505437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,10240,0.041111111640930176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,10240,0.04916622241338094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,7168,0.014815110299322339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,8192,0.034097777472601995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,8192,0.04121066795455085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,7168,0.03029511041111416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,7168,0.015824889143308003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,6144,0.013447110851605734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,12288,0.024734222226672705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,6144,0.014512000812424554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,65536,0.2727564440833198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,5120,0.010975999964608086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,64,0.00349866681628757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,5120,0.02310400042268965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,12288,0.02071466710832384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,4096,0.010269333091047075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,7168,0.036779556009504534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,4096,0.019353777170181274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,4096,0.024307555622524683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,4096,0.011758222348160215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,6144,0.031641778018739485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,8192,0.017087111870447796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,3584,0.017720000611411203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,5120,0.013131555583741931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,3072,0.007619555625650618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,5120,0.028095109595192805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,3072,0.010752889017264048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,3072,0.0158915552828047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,2560,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,3072,0.01958666741847992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,6144,0.02643377747800615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,2560,0.01036977767944336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,2560,0.01443911095460256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,2560,0.017640888690948486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,2048,0.0124195557501581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,1536,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,3584,0.021902221772405837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,1536,0.01053955571519004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,1536,0.013501333693663279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,1536,0.009076444639099969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,2048,0.005923555543025334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,1024,0.00868000007337994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,1024,0.008718222379684448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,3584,0.011077333655622272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,1024,0.011779555843936073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,2048,0.015499555402331881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,512,0.003724444243643019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,768,0.008354666332403818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,768,0.007687999970383114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,768,0.010429333481523724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,512,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,512,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,3584,0.008019555773999956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,512,0.010073777702119615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,128,0.0032222221295038858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,256,0.007364444434642792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,256,0.00868088917599784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,768,0.003921777837806278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,2048,0.009755555954244401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,128,0.005771555420425203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,32,0.0032239999208185407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,128,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,65536,0.06212888823615181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,64,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,32,0.005643555687533484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,1024,0.004496888981925117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,51200,0.05018044511477152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,65536,0.20629689428541395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,256,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,65536,0.27083643277486164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,16384,0.016199999385409884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,128,0.0075502221783002215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,51200,0.21351022190517852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,16384,0.054189334313074745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,256,0.0035422220826148987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,12288,0.013459555804729462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,65536,0.11679466565450032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,12288,0.041124443213144936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,10240,0.012275555895434486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,51200,0.09379733271068996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,51200,0.16067911518944636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,10240,0.03501244386037191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,10240,0.022376888328128394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,10240,0.0490515563223097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,16384,0.03198844525549147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,16384,0.07402311431037055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,7168,0.010591111249393888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,8192,0.017255110873116385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,8192,0.04042311178313361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,12288,0.05686133437686496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,7168,0.026008889079093933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,64,0.00300444434914324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,7168,0.03578044308556451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,6144,0.01459200017982059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,8192,0.011423110961914062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,5120,0.009759999811649323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,6144,0.02314666741424137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,6144,0.03174933460023668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,5120,0.013223110801643796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,12288,0.02478222217824724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,6144,0.012506666282812754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,4096,0.017102221647898357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,3584,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,4096,0.011766222616036734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,4096,0.02375999920898014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,3584,0.011361777782440186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,3072,0.007290666302045186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,8192,0.02892177634769016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,4096,0.00866844422287411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,3072,0.014182221558358936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,7168,0.01587199999226464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,2560,0.006304889089531369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,2560,0.012773333324326409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,2560,0.017589333984586928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,3584,0.015581332974963717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,3584,0.021670222282409668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,5120,0.02786133355564541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,2048,0.011071999867757162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,3072,0.010710222025712332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,1536,0.005031110925806893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,2048,0.00980888886584176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,2048,0.005623999983072281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,1536,0.009696000152164036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,5120,0.01995377739270528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,1536,0.01347555551264021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,1024,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,3072,0.019270223047998216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,1024,0.011735111474990845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,1024,0.008405333591832055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,768,0.007303111255168915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,2560,0.010208889014191097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,768,0.010391111175219217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,512,0.003576000117593341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,1024,0.00435288871328036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,1536,0.00905866672595342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,256,0.0032542221662071017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,512,0.007668444679843054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,512,0.006309333360857434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,512,0.009684444301658207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,2048,0.015508444772826301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,128,0.003192000091075897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,256,0.005640888793600931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,64,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,256,0.008678221868144141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,128,0.007352888584136963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,128,0.00526577772365676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,128,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,64,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,768,0.008028444316652086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,65536,0.20912444591522217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,256,0.007489778101444244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,16384,0.019753777318530612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,12288,0.011476444701353708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,32,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,10240,0.010739555789364709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,16384,0.053708444039026894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,8192,0.009702222214804756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,32,0.0053191110491752625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,7168,0.009409777820110321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,10240,0.03436977664629618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,6144,0.009012444151772393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,8192,0.028789334826999243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,5120,0.008343111309740279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,768,0.003919110943873723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,4096,0.007660444411966536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,6144,0.022658665974934895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,3584,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,5120,0.01997155613369412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,3072,0.006637333167923822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,65536,0.058724443117777504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,4096,0.017101332545280457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,2560,0.005657777604129579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,51200,0.04785155587726169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,3072,0.014131555954615274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,1536,0.004567999806669024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,2048,0.01110311100880305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,1024,0.0042204442951414315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,7168,0.025775111383861963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,768,0.003921777837806278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,51200,0.16018311182657877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,512,0.003206222214632564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,1024,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,256,0.0032355555643637977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,768,0.007348444726732042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,12288,0.04125688804520501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,512,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,64,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,256,0.005662222289376789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,2560,0.012632889052232107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,128,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,65536,0.05700711409250895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,64,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,51200,0.04605422086185879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,32,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,16384,0.012483555409643384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,65536,0.20912888314988878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,12288,0.01036266651418474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,128,0.002891555635465516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,51200,0.16011999713049993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,10240,0.00925866680012809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,8192,0.008542222281297049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,32,0.002903999967707528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,12288,0.041272001134024724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,7168,0.008218666745556725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,6144,0.007695111135641734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,8192,0.028586665789286297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,5120,0.007359999749395583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,7168,0.02555999987655216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,4096,0.00831200016869439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,6144,0.022856000396940444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,3584,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,5120,0.0199937770764033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,3072,0.006653333289755716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,3584,0.015526221858130561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,4096,0.016937777400016785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,2560,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,2048,0.005249777601824866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,3584,0.01568266749382019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,3072,0.014200000299347771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,10240,0.034824000464545354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,2560,0.012583110895421771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,1024,0.004211555752489301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,2048,0.011247110863526663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,768,0.0035706667436493766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,1024,0.008031999899281396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,512,0.0032435556252797446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,768,0.007359111474619971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,2048,0.0053182223604785065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,16384,0.054049776660071484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,128,0.0028782222006056043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,512,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,64,0.0028355556229750314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,256,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,1536,0.004550222307443619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,32,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,128,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,1536,0.009409777820110321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,64,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,32,0.005032000028424793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,1536,0.009738666315873465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,256,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,16384,1.0062435997856987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,12288,2.3545199500189886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,12288,0.7371270921495227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,16384,3.3173431820339627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,16384,3.3295004102918835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,12288,2.4895990159776473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,10240,0.6264275444878472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,51200,3.6447660658094616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,10240,1.9916523827446833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,8192,1.6154996024237738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,8192,0.5150791274176704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,10240,2.024479971991645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,8192,1.5287226570977106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,7168,0.4722275733947754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,7168,1.4905057483249242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,7168,1.2906746334499783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,6144,0.38948623339335126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,6144,1.416177749633789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,6144,1.0978115929497612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,5120,0.33363111813863117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,4096,0.7779048813713921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,4096,0.2771093315548367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,5120,0.9341146681043836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,4096,0.7446604304843478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,3584,0.2461493280198839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,3072,0.5149493217468262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,5120,0.9864266713460287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,3584,0.6419351365831163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,3072,0.21721243858337402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,3072,0.5420791308085123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,2560,0.43708088662889266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,2560,0.18994222746955025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,2048,0.40172621938917374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,3584,0.6739075448777941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,2560,0.4639537599351671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,2048,0.1622355514102512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,2048,0.3805600007375081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,1536,0.1267146666844686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,51200,10.097835964626736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,1024,0.10302044285668267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,768,0.17505510648091635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,1024,0.22804088062710234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,768,0.09072888559765285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,512,0.08095288938946195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,512,0.1565422217051188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,1536,0.2831137710147434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,1024,0.22956355412801108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,256,0.0730817781554328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,1536,0.33276889059278697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,256,0.14040799935658774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,64,0.06124089161554972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,512,0.1258257759941949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,128,0.1325955523384942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,32,0.06395200226042005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,64,0.13384089205000135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,768,0.18325511614481607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,32,0.134624891810947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,128,0.06602133644951715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,128,0.07044355736838447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,256,0.09727199872334798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,51200,13.704646640353731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,16384,2.4886205461290145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,16384,2.6159600151909723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,51200,2.9368871053059897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,16384,0.7639590899149576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,12288,1.7378418180677626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,65536,3.634507497151693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,12288,1.9414747026231554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,12288,0.5750826729668511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,51200,7.53025139702691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,10240,0.48983022901746964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,10240,1.5204586452907984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,8192,0.3901226785447862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,8192,1.4227475060356989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,65536,9.674504597981771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,8192,1.2080444759792752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,7168,1.108327971564399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,7168,1.038657824198405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,10240,1.444535149468316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,6144,0.30931909879048664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,6144,0.8697137832641602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,51200,10.857535468207466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,5120,0.2614702118767632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,6144,1.0046026441786025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,5120,0.7410382164849175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,4096,0.649028460184733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,3584,0.5774115456475152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,3584,0.19523821936713326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,4096,0.5521395471360948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,4096,0.21678666273752847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,3072,0.16962222258249918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,3072,0.5569929016960992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,2560,0.41089778476291233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,2560,0.17549688286251494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,2048,0.326507568359375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,7168,0.4271715482076009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,5120,0.7221066686842176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,3584,0.4934817949930827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,2048,0.14384533299340144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,65536,14.048269483778212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,2048,0.33694667286343044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,1024,0.17236710919274223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,1536,0.10122666756312053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,3072,0.4473759863111708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,768,0.1365777783923679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,1024,0.19018844763437906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,512,0.1026817758878072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,1536,0.2371306684282091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,2560,0.3645635445912679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,256,0.07346755266189575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,512,0.06551200151443481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,128,0.05307555529806349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,256,0.058825777636633984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,256,0.11097511317994858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,64,0.04930311110284594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,1024,0.07960089047749837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,32,0.05165511038568285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,768,0.07143288850784302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,64,0.10564888848198785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,32,0.10658399926291572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,512,0.12377244896358913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,1536,0.26546311378479004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,128,0.0566177765528361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,128,0.10486044486363728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,768,0.14299644364251032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,16384,0.6845368809170194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,65536,1.106471061706543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,51200,0.8948311275906033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,16384,0.25433778762817383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,16384,0.7999004258049859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,12288,0.1930382251739502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,12288,0.7661333084106445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,10240,0.5072844293382432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,12288,0.7545555432637533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,8192,0.40211110644870335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,10240,0.483367125193278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,8192,0.1347528828514947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,7168,0.30468800332811147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,51200,3.603570726182726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,51200,2.4124488830566406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,65536,3.0831947326660156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,65536,4.350704193115234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,6144,0.30621245172288686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,10240,0.164520886209276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,5120,0.25333865483601886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,6144,0.10782311360041301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,6144,0.2853422164916992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,5120,0.08982844485176934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,4096,0.19542222552829316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,5120,0.2676764329274495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,4096,0.0765297810236613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,8192,0.38324888547261554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,3072,0.15083732869890001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,3584,0.16689244906107584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,7168,0.12004266844855414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,3072,0.06038577689064873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,2560,0.12063022454579671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,7168,0.33897156185574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,2048,0.09325510925716823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,2560,0.12569422192043728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,2048,0.046168890264299184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,1536,0.07493688662846883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,3584,0.17550399568345812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,1024,0.05236533284187317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,3584,0.06854489114549425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,1536,0.08338755369186401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,4096,0.1884151167339749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,768,0.04120622078577677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,1024,0.03107644451989068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,1024,0.06280355321036445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,768,0.028983109527164038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,2560,0.05335466729270088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,768,0.050760888391070895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,512,0.026685333914226953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,512,0.04377422067854139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,2048,0.1035075585047404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,128,0.020084444019529555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,256,0.03938044442070855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,1536,0.03773511118359036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,64,0.01754844519827101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,128,0.02342133389578925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,128,0.037179556157853864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,32,0.018193778064515855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,3072,0.14611821704440647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,64,0.037112000915739275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,32,0.037548445992999606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,256,0.024380443824662104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,512,0.03447377681732178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,65536,0.8639520009358724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,256,0.024550222688251074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,16384,0.6698915693495008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,65536,2.4441030290391708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,51200,0.6936488681369357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,16384,0.6060586505466038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,16384,0.2157848940955268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,12288,0.1510817739698622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,10240,0.3487164444393582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,10240,0.12858045101165771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,12288,0.5525493092007107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,51200,2.6029040018717446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,10240,0.3686773247188992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,8192,0.10628000232908462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,51200,2.1191467709011502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,8192,0.2853848934173584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,7168,0.09515111313925849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,6144,0.2301271226671007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,7168,0.25137421819898814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,12288,0.43715111414591473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,5120,0.18948800033993193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,6144,0.2160408894220988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,4096,0.15066311094495985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,5120,0.18125066492292616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,3584,0.11596443918016221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,4096,0.14510133531358507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,7168,0.27737779087490505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,3072,0.11416178279452854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,3584,0.052186667919158936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,3584,0.12905422846476236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,3072,0.046295109722349376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,2560,0.08449688884947036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,3072,0.11288000477684869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,65536,3.156679153442383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,2560,0.09717688957850139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,4096,0.05812178055445353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,1536,0.05888977977964613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,2048,0.08003111018074884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,1536,0.02927288744184706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,1536,0.06377244657940336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,6144,0.08334133360120986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,1024,0.02403822210099962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,1024,0.04832888974083794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,2048,0.035756445593304105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,768,0.03885066509246826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,512,0.024129778146743774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,1024,0.04327022367053562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,512,0.03211288981967502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,512,0.020271999968422782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,2560,0.04115733173158433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,768,0.0322088897228241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,128,0.01591466698381636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,256,0.018571555614471436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,256,0.02902488907178243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,64,0.013434666726324292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,128,0.018272888329293992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,128,0.027848889430363972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,32,0.013711111413107978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,64,0.027220444546805486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,32,0.027211555176311072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,256,0.018938667244381376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,768,0.02199466692076789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,65536,0.7382666799757215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,65536,1.8912417093912761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,51200,1.786518308851454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,16384,0.4609644677903917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,51200,0.5887706544664171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,16384,0.1667546696133084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,12288,0.38367022408379453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,8192,0.2732497851053874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,16384,0.6542773246765137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,12288,0.1612702210744222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,10240,0.26452623473273384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,51200,2.200818591647678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,12288,0.3615662256876628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,10240,0.11507644918229844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,8192,0.21317244900597465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,10240,0.3122026655409071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,7168,0.23002934455871582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,7168,0.07885689205593534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,65536,2.6019742753770614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,6144,0.197035551071167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,7168,0.21463021967146131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,5120,0.1677795516120063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,6144,0.18513333797454834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,5120,0.0624106658829583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,5120,0.1545999977323744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,8192,0.088736891746521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,4096,0.05044088761011759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,8192,0.25866577360365123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,4096,0.12506133980221218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,3584,0.09888088703155518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,2048,0.07371199793285794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,3584,0.044922666417227856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,3584,0.11023022068871392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,3072,0.039889779355790876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,2560,0.07668089204364352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,3072,0.09631733099619548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,4096,0.12936977545420328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,2560,0.082424885696835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,2048,0.0303457776705424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,1536,0.04943822158707512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,5120,0.07077778047985502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,6144,0.06894222233030531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,1536,0.024918221765094336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,1024,0.036983999941084124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,1536,0.0542702211274041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,768,0.0270506673389011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,1024,0.021125333176718816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,1024,0.041486223538716636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,512,0.026704000102149114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,768,0.033871110942628645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,2048,0.06789333290523954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,512,0.01759466694460975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,3072,0.0823813345697191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,512,0.0276826669772466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,128,0.013829333086808523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,256,0.0162808895111084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,64,0.012087110843923358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,256,0.024777778320842322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,128,0.01612977849112617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,32,0.012122666670216454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,2048,0.060919112629360624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,64,0.02367377777894338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,32,0.023476445012622412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,768,0.019311999281247456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,256,0.02164711058139801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,65536,0.6087910864088271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,128,0.023744889431529578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,2560,0.03485511077774896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,65536,1.5109039942423503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,16384,0.4513395627339681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,16384,0.14056266678704157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,12288,0.2907662126753065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,16384,0.4050942262013753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,12288,0.10783644517262776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,10240,0.271708435482449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,65536,1.9948889414469402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,12288,0.4275537861718072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,10240,0.09143111440870498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,8192,0.20329956213633218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,8192,0.07551022370656331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,8192,0.2047546704610189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,7168,0.06738844182756212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,6144,0.1573208835389879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,7168,0.1749288903342353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,5120,0.12190755208333333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,6144,0.057576000690460205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,6144,0.15038843949635824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,5120,0.049472000863817006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,4096,0.11513422595130073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,10240,0.2451039950052897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,7168,0.19212178389231363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,4096,0.10229688882827759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,3584,0.03696889016363356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,3072,0.078366219997406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,5120,0.12673955493503147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,51200,0.4722657733493381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,3072,0.03320622113015916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,2560,0.06877599822150336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,2560,0.030121776792738173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,3584,0.09106489022572835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,2048,0.0560471084382799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,51200,1.5759076012505426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,1536,0.04337866769896614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,2048,0.026131555438041687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,2048,0.05640800131691826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,1024,0.02926666537920634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,1536,0.021540444758203294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,51200,1.1494151221381295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,1536,0.04530222217241923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,1024,0.017954667409261067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,1024,0.034985777404573225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,512,0.017088888419999015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,768,0.02858577834235297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,2560,0.0683528913391961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,256,0.014044443766276041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,512,0.015262222952312894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,256,0.014426666829321118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,3072,0.07946933640374078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,3584,0.097634666495853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,768,0.02204533252451155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,64,0.010410666465759277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,128,0.014296889305114746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,128,0.020006222857369315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,768,0.016639111770523917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,64,0.0199617776605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,32,0.01974933346112569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,128,0.012083555261294046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,4096,0.041215999258889094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,32,0.010680889089902243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,65536,0.5140862464904785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,512,0.023401778605249193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,256,0.02067377832200792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,16384,0.34178490108913845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,51200,1.1931493547227647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,51200,0.40144978629218203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,16384,0.12405067019992405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,12288,0.3228026760949029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,16384,0.3509191142188178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,51200,1.3543270958794489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,10240,0.2704177697499593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,12288,0.29021067089504665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,10240,0.08214222060309516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,10240,0.2180640035205417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,8192,0.06628888845443726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,8192,0.17785067028469512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,65536,1.8088550567626953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,6144,0.13713155852423772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,7168,0.0583546691470676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,7168,0.1552248928281996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,6144,0.05195111036300659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,65536,1.7086861928304036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,6144,0.13364711072709826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,4096,0.0899839997291565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,5120,0.044008000029457935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,8192,0.2006657785839505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,3584,0.08417066600587632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,4096,0.036665777365366616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,3072,0.07080088721381293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,3584,0.03308533297644721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,7168,0.12683644559648302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,3584,0.08129777510960896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,2560,0.055678221914503306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,3072,0.03018310997221205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,2048,0.04439466529422336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,2560,0.02737155556678772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,2560,0.06076622009277344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,1536,0.034697777695126004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,2048,0.023795555035273235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,2048,0.050760888391070895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,1024,0.023353776997990076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,5120,0.11377244525485569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,1536,0.04099555479155646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,4096,0.0915644433763292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,1024,0.016906667086813185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,1024,0.03130933311250474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,512,0.014747555057207743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,768,0.015397333436542086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,3072,0.07054311037063599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,5120,0.09641066524717544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,512,0.021163556310865615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,256,0.013345777988433838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,256,0.018647111124462552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,128,0.010384889112578498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,128,0.013456888496875763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,64,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,128,0.018182223041852314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,32,0.009115555220180089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,768,0.02591288917594486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,12288,0.09471822447246975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,64,0.01791555517249637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,32,0.017887999614079792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,1536,0.02015644477473365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,768,0.019101333287027147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,65536,0.4880808724297418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,51200,0.820308420393202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,65536,1.188370704650879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,256,0.011744889120260874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,51200,0.38956088489956325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,16384,0.136172440316942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,12288,0.23351287841796875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,16384,0.32949688699510365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,65536,1.658000840081109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,10240,0.1991057793299357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,51200,1.1145298216078017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,12288,0.2362719906700982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,8192,0.15708711412217882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,10240,0.0826044413778517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,7168,0.13831378353966609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,10240,0.19496622350480822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,8192,0.15751643975575766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,6144,0.10118755367067124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,16384,0.31257332695855033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,7168,0.1365777783923679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,5120,0.10413777828216553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,6144,0.11832000149620904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,5120,0.09980089134640163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,5120,0.03862399856249491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,4096,0.03300977746645609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,8192,0.05990933047400581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,4096,0.07963999774720934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,3584,0.07268799675835504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,512,0.01422933406300015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,7168,0.059855110115475126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,3584,0.02994222111172146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,3584,0.07142933209737141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,3072,0.02770044406255086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,6144,0.045442667272355824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,12288,0.08726933267381455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,2560,0.05342488818698459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,2048,0.044976890087127686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,3072,0.05950399902131823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,2048,0.021354667014545862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,2560,0.047071999973720975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,3072,0.06217155853907267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,1024,0.022467555271254644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,1536,0.03654044535424974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,4096,0.07510222329033746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,768,0.017121778594122995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,1024,0.02787289023399353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,768,0.014356444279352823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,768,0.022999111149046157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,512,0.014143110977278816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,1536,0.03130577670203315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,256,0.01057866629627016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,512,0.013378666506873237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,1536,0.017829333742459614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,128,0.009219555391205682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,256,0.01688888834582435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,64,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,1024,0.01531377765867445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,128,0.012537777423858643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,128,0.016184889607959323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,64,0.016189333465364244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,32,0.016184889607959323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,512,0.01906488835811615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,256,0.012421333127551608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,2048,0.03963911202218797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,65536,0.42639909850226504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,65536,1.124966197543674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,16384,0.29140710830688477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,51200,0.3353128963046604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,51200,0.7930169105529785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,65536,1.253164397345649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,16384,0.1140471167034573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,12288,0.21220889356401232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,12288,0.07680622074339125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,12288,0.19285244411892363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,10240,0.06475022104051378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,8192,0.11516622702280681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,32,0.00832799987660514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,7168,0.11913866466946071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,8192,0.05146311057938469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,8192,0.13428266843159994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,7168,0.046097778611712985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,16384,0.26629334025912815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,7168,0.11754488945007324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,10240,0.17251288890838623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,6144,0.04134755664401584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,2560,0.02457777824666765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,5120,0.0350888901286655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,4096,0.06285333633422852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,5120,0.08583555618921916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,3584,0.06164533562130398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,4096,0.06907288895712958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,3584,0.026879111925760906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,3584,0.0613466633690728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,3072,0.05053155620892843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,51200,0.9737013710869683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,2560,0.04065510961744521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,3072,0.02460711035463545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,3072,0.05378577775425381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,10240,0.16382843918270534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,2560,0.02204088866710663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,2560,0.04633333285649618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,4096,0.029719998439153034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,2048,0.019158222609096102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,2048,0.04012711180580987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,1536,0.016757332616382174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,6144,0.10547377665837605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,1536,0.03176622258292304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,768,0.015602666470739575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,5120,0.079221334722307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,1024,0.024120888776249353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,6144,0.10096889071994358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,512,0.012014221813943652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,768,0.013392888837390475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,2048,0.033117334047953285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,256,0.012085333466529846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,512,0.012333333492279053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,128,0.008036444584528605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,1536,0.0262097782558865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,256,0.014869333969222175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,64,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,128,0.014446222119861178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,32,0.007711110843552484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,1024,0.014330666926172046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,64,0.014699555105633207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,32,0.014142221874660917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,768,0.020391111572583515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,512,0.016916443904240925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,256,0.011744000017642975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,65536,0.7437546518113879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,65536,0.3620399898952908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,1024,0.01958577831586202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,51200,0.5747173097398546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,16384,0.2166879971822103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,51200,0.28730755382113987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,12288,0.16545422871907553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,65536,1.1092408498128254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,16384,0.2279102272457547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,12288,0.06851110855738322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,10240,0.1341493394639757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,51200,0.7810399797227648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,12288,0.16461777687072754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,8192,0.10751733514997695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,10240,0.05573866764704386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,7168,0.09904355472988552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,8192,0.11034311188591851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,128,0.011796444654464722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,6144,0.08263999885982938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,16384,0.08486221896277533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,6144,0.03485688898298476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,5120,0.06583822435802884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,6144,0.0850915577676561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,4096,0.05257688959439596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,5120,0.031109333038330078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,5120,0.07180088758468628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,3584,0.04913599954711067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,4096,0.05776088767581516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,7168,0.04342933495839437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,3584,0.023668444818920557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,3584,0.05206222335497538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,3072,0.02132533325089349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,3072,0.04556177722083198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,2560,0.0344835552904341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,2048,0.027102222045262653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,2560,0.01936177743805779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,4096,0.02621777852376302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,1536,0.023025777604844835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,2048,0.016946666770511203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,2048,0.03341689043574863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,1024,0.01719288859102461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,1536,0.014581332604090372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,1536,0.02683999968899621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,768,0.014492443866199918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,1024,0.020638222495714825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,768,0.012248000337017907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,512,0.011400889191362592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,768,0.017639999588330586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,10240,0.13949778344896105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,256,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,512,0.011408888631396823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,512,0.014576888746685453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,7168,0.09832266966501872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,128,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,256,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,256,0.013109332985348172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,64,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,128,0.01073688848151101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,128,0.012498666842778524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,64,0.012429333395428128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,32,0.012460444536474017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,8192,0.04355288876427544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,65536,0.5439662403530544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,3072,0.03901777664820353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,51200,0.5630044407314724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,32,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,51200,0.23396266831292045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,16384,0.1778115563922458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,12288,0.13501866658528647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,16384,0.07173155413733588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,16384,0.19490578439500597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,12288,0.05520622266663445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,12288,0.15776888529459634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,10240,0.04855111241340637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,8192,0.09707288609610663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,51200,0.7256106800503201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,2560,0.03931822379430135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,8192,0.10058044724994236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,7168,0.0853946672545539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,65536,0.3016800085703532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,7168,0.034725334909227155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,6144,0.0702435572942098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,1024,0.012883555557992725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,6144,0.030553778012593586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,6144,0.07643822166654798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,5120,0.026924444569481745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,5120,0.06490221950742933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,4096,0.045500444041358105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,8192,0.03791466686460707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,3584,0.04090577695104811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,4096,0.02335111134582096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,4096,0.05236622360017565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,7168,0.0884480012787713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,3072,0.033936000532574125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,10240,0.12514755460951063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,3584,0.047239998976389565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,65536,0.9513422648111979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,3072,0.04155022237035964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,2560,0.017271111408869427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,2048,0.024307555622524683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,2560,0.036207109689712524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,1536,0.019454222586419847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,2048,0.015484443969196744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,10240,0.11674311425950791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,2048,0.03091111116939121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,1536,0.013777777552604675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,1536,0.024511110451486375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,3584,0.021592888567182753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,768,0.011763555308183035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,1024,0.012098666694429187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,3072,0.019170666734377544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,512,0.009359999663299983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,768,0.011407111254003314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,768,0.01645599967903561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,256,0.007515555454625024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,512,0.010716444088353051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,512,0.01349244432316886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,256,0.010079999764760336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,256,0.012127111355463663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,128,0.010069333016872406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,128,0.011743110915025076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,64,0.005986666513813867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,32,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,64,0.011768889096048145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,128,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,65536,0.4860853089226617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,2560,0.02924355533387926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,65536,0.274637328253852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,51200,0.4658942222595215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,65536,0.7936204274495443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,16384,0.15366044309404162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,51200,0.21747733487023246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,51200,0.6304453214009603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,12288,0.11465511057111953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,16384,0.1755982240041097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,5120,0.05572977993223402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,12288,0.05407822132110596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,12288,0.13195466995239258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,10240,0.047114666965272695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,8192,0.07687911060121325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,10240,0.1097973320219252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,7168,0.06563378042644925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,8192,0.08960622549057007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,16384,0.0773093303044637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,7168,0.03352533446417914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,7168,0.07827199829949273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,1024,0.019147555033365887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,5120,0.048017776674694486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,6144,0.06854844755596586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,5120,0.025806221697065566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,32,0.011577777564525604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,4096,0.040652443965276085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,6144,0.05903289053175184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,4096,0.04689955711364746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,6144,0.029432889488008287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,3584,0.04210755560133192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,8192,0.03661866651640998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,3072,0.028889778587553237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,1024,0.014313777287801107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,2560,0.025507556067572698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,3072,0.03741511040263706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,4096,0.02217955556180742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,10240,0.09005333317650689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,2048,0.021232000655598108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,3584,0.03341689043574863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,1536,0.01724177764521705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,2560,0.0327413347032335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,2048,0.015057777365048727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,1024,0.013342222405804528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,2048,0.027624888552559748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,1536,0.01348000019788742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,3584,0.020351111888885498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,768,0.011150221857759686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,1024,0.011752000285519494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,1024,0.01718488832314809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,5120,0.05766044722663032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,768,0.014926221635606555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,256,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,512,0.012470222181744046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,2560,0.01688088807794783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,128,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,256,0.009758222434255812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,256,0.011120888921949597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,64,0.005578666511509154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,128,0.009702222214804756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,128,0.010728889041476779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,512,0.010057777994208867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,768,0.011074666347768573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,32,0.010750222537252637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,512,0.010438222024175854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,65536,0.4560142093234592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,65536,0.23924620946248373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,51200,0.4231662220425076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,16384,0.1379635598924425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,51200,0.19650844732920328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,64,0.01072622256146537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,3072,0.018595554762416415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,16384,0.06039110819498698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,12288,0.0920284456676907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,16384,0.15807467036777073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,12288,0.046952890025244824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,12288,0.11352533764309353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,10240,0.04107644491725498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,8192,0.07562577724456787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,32,0.005595555735958948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,8192,0.0788044465912713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,7168,0.06524266799290974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,51200,0.5466817749871148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,7168,0.02978666623433431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,10240,0.0888248880704244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,7168,0.06897688574261136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,10240,0.09608978033065796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,6144,0.05960444609324137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,5120,0.045122666491402515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,65536,0.6725777520073785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,4096,0.035841776265038386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,5120,0.023737778266270954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,5120,0.050675554407967456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,4096,0.020655110478401184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,4096,0.042135109504063926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,6144,0.057814220587412514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,3584,0.01848444508181678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,6144,0.02723466687732273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,8192,0.03442755672666762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,3072,0.03331911232736375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,3584,0.03264710969395108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,2048,0.020409777760505676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,2560,0.016135111451148987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,3584,0.03747999999258254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,2048,0.0148062225845125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,1536,0.016573333077960543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,1536,0.022207111120224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,2560,0.023844444089465674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,1536,0.019819556011093985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,1024,0.010732444624106089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,768,0.010187555518415239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,2560,0.02882044514020284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,768,0.010024888647927178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,3072,0.02998133169280158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,512,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,2048,0.02444177700413598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,256,0.006686222222116258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,512,0.009726222190592024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,3072,0.016547555724779766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,1024,0.012435555458068848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,256,0.009023110899660323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,256,0.010201777848932479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,64,0.005236444373925527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,128,0.009015111459626092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,1024,0.01551644504070282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,32,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,64,0.009749333063761393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,768,0.013501333693663279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,65536,0.41317955652872723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,1536,0.01220355596807268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,128,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,51200,0.3277031050788032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,16384,0.11060355769263373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,51200,0.16670578055911592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,51200,0.4127262168460422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,16384,0.0563653310139974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,12288,0.08952889177534316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,12288,0.04296355446179708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,512,0.011197333534558615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,10240,0.07200088765886095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,10240,0.03746577766206529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,65536,0.5218319892883301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,8192,0.056220445368025035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,8192,0.029723554849624634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,8192,0.06807466348012288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,12288,0.09850844409730698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,65536,0.20992533365885416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,6144,0.03869600097338358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,7168,0.02701955537001292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,10240,0.08188622527652316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,6144,0.023984000086784363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,5120,0.035587555832333036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,6144,0.051813334226608276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,4096,0.026352889007992212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,5120,0.02109688851568434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,5120,0.04355644517474704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,4096,0.01815200017558204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,4096,0.035974221097098455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,7168,0.05965422259436714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,3072,0.020999999509917364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,3584,0.01686311099264357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,3584,0.032279110617107816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,2560,0.01815111107296414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,128,0.009697777529557547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,3072,0.02863022353914049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,2048,0.01591466698381636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,2560,0.024622221787770588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,16384,0.13026755385928684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,1536,0.012772444221708508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,2048,0.01315111087428199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,3584,0.0236835562520557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,1024,0.01038755559258991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,1536,0.011759111450778114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,1536,0.01756266587310367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,768,0.008711111214425828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,1024,0.010402667025725046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,1024,0.01370133293999566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,3072,0.015654222832785714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,768,0.009720889230569204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,768,0.011745777395036487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,512,0.009366222553782994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,32,0.009726222190592024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,512,0.010057777994208867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,2048,0.02107644412252638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,256,0.009004444711738164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,128,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,256,0.009105777574910058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,64,0.004605333424276776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,128,0.008990222381220924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,32,0.005060444275538127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,512,0.007330666813585494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,64,0.008543111383914948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,2560,0.014479110638300577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,32,0.00870666652917862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,7168,0.04858488837877909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,256,0.005620444400442972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,65536,0.3395422299702962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,16384,0.08595555358462864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,51200,0.13204977247450086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,65536,0.4172942108578152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,51200,0.3347280025482178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,128,0.008678221868144141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,12288,0.03499822152985467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,12288,0.08074133263693915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,10240,0.05381955703099569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,8192,0.04380711250834995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,10240,0.031090666850407917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,51200,0.22490843137105307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,10240,0.0678115553326077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,7168,0.03462844424777561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,8192,0.055664002895355225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,6144,0.03198400139808655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,7168,0.022138666775491502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,16384,0.11294933160146077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,7168,0.04922222097714742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,5120,0.02730400032467312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,6144,0.01981244484583537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,4096,0.022636445032225713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,5120,0.01761511133776771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,5120,0.03676177726851569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,65536,0.16496266259087458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,8192,0.024477332830429077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,3584,0.014476444986131458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,3072,0.01791288952032725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,6144,0.04295733239915636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,3072,0.02443466583887736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,2560,0.01586666703224182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,4096,0.015498666299713982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,2560,0.021196444829305012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,2560,0.012443555725945367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,12288,0.06719288561079237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,3584,0.027172444595230952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,1536,0.011258666714032492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,2048,0.018057778477668762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,1536,0.010403555300500657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,1024,0.009024000002278222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,4096,0.030745777818891738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,1024,0.01203644441233741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,16384,0.04469777809249031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,2048,0.011329777538776398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,768,0.010407110883129967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,768,0.009087111386987898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,512,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,3072,0.01349955548842748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,256,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,512,0.008674666285514832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,512,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,128,0.004584889031118817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,1024,0.009388444324334463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,768,0.007736889024575551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,64,0.004229333251714706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,128,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,128,0.00776800016562144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,32,0.004600000050332812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,65536,0.2068915499581231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,32,0.007701333198282454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,3584,0.02049333353837331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,65536,0.1576639943652683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,51200,0.17141511705186632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,51200,0.12504888905419245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,51200,0.25472977426317006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,16384,0.06157333321041531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,64,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,1536,0.015115555789735583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,12288,0.04453777935769823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,16384,0.08442222409778172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,10240,0.03714488943417867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,256,0.008353778057628209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,12288,0.0632515549659729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,2048,0.013502221968438892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,8192,0.029130667448043823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,10240,0.029232889413833618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,7168,0.027264000640975103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,8192,0.022767111659049988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,8192,0.04394222299257914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,7168,0.02070399953259362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,6144,0.022801778382725183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,7168,0.039170665873421565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,16384,0.042055110136667885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,6144,0.018608000543382432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,6144,0.03466755482885573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,12288,0.032631112469567194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,4096,0.015987555185953777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,5120,0.016748444901572335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,10240,0.05353866683112251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,3584,0.014525334040323893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,5120,0.019703111714786954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,3584,0.013711111413107978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,3072,0.012888888518015543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,256,0.008044444852405125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,4096,0.014575110541449653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,3072,0.012770666844315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,4096,0.024449777272012498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,2560,0.01178311142656538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,2560,0.017432888348897297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,3584,0.022129777404997084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,2048,0.010641777680979835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,1536,0.008319999608728621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,65536,0.3219368987613254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,1024,0.006746666712893381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,1536,0.010033778018421596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,1536,0.012545777691735161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,768,0.006248000181383557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,1024,0.010188444621033138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,2048,0.009530666801664565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,768,0.008807111117574904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,512,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,768,0.009111111362775167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,256,0.004576888763242298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,512,0.00814844419558843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,2048,0.015116444892353482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,512,0.007711999946170383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,128,0.003887999802827835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,256,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,256,0.007361777954631382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,128,0.008032889001899296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,3072,0.019959110352728102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,128,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,1024,0.008997333546479544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,64,0.00674222202764617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,65536,0.15867555141448975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,32,0.006989333364698622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,51200,0.12038221624162461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,65536,0.15494667159186468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,65536,0.28231467141045463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,5120,0.029456890291637842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,51200,0.2210435602400038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,64,0.003718222181002299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,12288,0.036335110664367676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,16384,0.04147911071777344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,16384,0.073381331231859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,32,0.003899555653333664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,12288,0.03245333168241713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,8192,0.023364444573720295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,10240,0.0470124449994829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,8192,0.022568888134426538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,16384,0.04631111025810242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,7168,0.020482665962643094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,51200,0.12279378043280707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,6144,0.017587555779351127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,7168,0.02068622244728936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,7168,0.03470844361517165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,2560,0.011300444602966309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,5120,0.015054222610261707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,6144,0.0307279990779029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,4096,0.012919999659061432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,5120,0.02600266701645321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,4096,0.02185955478085412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,3584,0.0118222220076455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,3584,0.019669334093729656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,10240,0.027570666538344488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,3072,0.011005333728260465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,3584,0.013839999834696451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,10240,0.028766221470303003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,2560,0.009711999860074785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,3072,0.012827555338541666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,3072,0.017761778500345018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,2048,0.008463111188676622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,2560,0.015504888362354703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,2048,0.010508444574144153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,1536,0.0074177781740824384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,6144,0.01844800015290578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,12288,0.05600533220503065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,1024,0.006236444330877728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,8192,0.03879377908176846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,1024,0.009379555781682333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,768,0.005582222094138463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,1024,0.00904799997806549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,2048,0.013501333693663279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,768,0.008681777450773451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,768,0.008352888955010308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,1536,0.009729777773221334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,512,0.00833155545923445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,4096,0.014545778433481852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,256,0.006882666713661618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,128,0.0038986665507157645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,512,0.004566222015354368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,128,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,128,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,5120,0.01661777827474806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,256,0.004237333519591225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,32,0.003879111260175705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,64,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,2560,0.011797332929240333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,65536,0.13542132907443577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,32,0.006329777754015393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,51200,0.11073688666025798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,65536,0.1122808853785197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,65536,0.23513333002726236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,16384,0.036166220903396606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,64,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,51200,0.09056355555852254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,512,0.007333333293596904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,16384,0.030232889784706965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,16384,0.061455110708872475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,10240,0.0217884447839525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,12288,0.04686399963166979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,256,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,8192,0.018159111340840656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,10240,0.021951999929216173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,10240,0.040285332335366145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,7168,0.016418667303191293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,8192,0.03324088785383437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,1536,0.01146933353609509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,7168,0.015635554989178974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,6144,0.014569777581426831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,12288,0.024123556084103052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,5120,0.012862222062216865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,6144,0.02624799973434872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,4096,0.011043555206722684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,5120,0.02236266599761115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,4096,0.011767110890812345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,51200,0.18836000230577257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,3584,0.009746666583749983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,7168,0.029482665989134047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,3072,0.008707555631796518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,3584,0.011092444260915121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,6144,0.01423822177780999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,3584,0.017218665944205392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,2560,0.007984889050324758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,3072,0.010439111126793755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,3072,0.01577422188387977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,2560,0.009992000129487779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,5120,0.01313066648112403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,2560,0.013842666314707862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,1536,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,2048,0.009386666946940953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,1024,0.004956444518433677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,1536,0.008998221821255153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,4096,0.019024888674418133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,12288,0.025772444076008264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,768,0.004584889031118817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,1024,0.008350222474998897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,512,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,768,0.007699555820888943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,2048,0.006897777732875612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,256,0.0038782221575578055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,512,0.006648889018429651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,8192,0.016876444220542908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,2048,0.01219466659757826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,1536,0.010378667049937779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,256,0.0059262220230367445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,64,0.0032035555276605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,128,0.007283555964628856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,768,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,64,0.005592888842026393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,65536,0.07440000110202365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,128,0.003552000141806073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,65536,0.19382754961649576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,51200,0.0610355536142985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,1024,0.00871288859181934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,51200,0.08983111381530762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,51200,0.15273243851131862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,32,0.005635555419656966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,16384,0.029723554849624634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,16384,0.051242666112052075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,256,0.007290666302045186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,12288,0.02360533343421088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,32,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,12288,0.03875733415285746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,10240,0.021659556362364028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,8192,0.013759999639458127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,10240,0.033075554503334895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,8192,0.01688888834582435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,7168,0.01256977766752243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,128,0.005658666706747479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,7168,0.024719999896155462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,6144,0.01092088884777493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,7168,0.01575999955336253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,512,0.007696000238259633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,6144,0.021758221917682227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,12288,0.01827022267712487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,5120,0.019155555301242404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,4096,0.008439111212889353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,65536,0.11176089445749919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,4096,0.01166488892502255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,4096,0.01628444426589542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,3584,0.007648000286685095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,16384,0.021866665946112737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,3072,0.006955555743641323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,3584,0.01108000013563368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,3584,0.014886221951908536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,8192,0.027503111296229895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,2560,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,3072,0.010552888943089379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,5120,0.013042666845851474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,2048,0.005620444400442972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,2560,0.010027555955780877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,2560,0.012145777543385824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,1536,0.00508622204264005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,2048,0.009348444640636444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,2048,0.01075466639465756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,6144,0.014175999495718213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,1024,0.004306666553020477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,5120,0.009745777481132084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,1536,0.00924711094962226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,1024,0.00832088871134652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,1024,0.00775022225247489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,512,0.003859555555714501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,768,0.008046222229798635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,768,0.00720177756415473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,3072,0.013597333596812354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,256,0.003560888684458203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,10240,0.01642311116059621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,512,0.007638221813572778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,128,0.0032471112079090546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,512,0.006118222243256039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,64,0.002965333354141977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,128,0.007151111132568783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,32,0.003157333367400699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,128,0.0052435555391841466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,1536,0.009004444711738164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,65536,0.06102933486302694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,64,0.0052800000541739995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,51200,0.04890577660666572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,65536,0.11127021577623154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,16384,0.014885332849290637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,51200,0.08976622422536214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,51200,0.14817333221435547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,256,0.007299555672539606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,16384,0.029338667790095013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,256,0.00562755556570159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,10240,0.012900444368521372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,12288,0.023239110906918842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,12288,0.03835555579927232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,768,0.0038986665507157645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,8192,0.012924444344308643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,10240,0.021632888250880774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,32,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,10240,0.03233866559134589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,8192,0.016586666305859882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,65536,0.1927973296907213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,7168,0.01551466683546702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,7168,0.024151111642519634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,6144,0.014177777700954013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,12288,0.014513777361975776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,5120,0.013081777426931592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,5120,0.018615111708641052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,4096,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,4096,0.011577777564525604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,3584,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,4096,0.016207999653286405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,6144,0.010374222364690568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,8192,0.026879111925760906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,3072,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,3584,0.011071999867757162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,2560,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,3072,0.010386666489972008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,16384,0.050372444921069674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,3072,0.013628444737858243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,2560,0.010040000081062317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,2560,0.012127111355463663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,1536,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,2048,0.009377777576446533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,2048,0.010761777559916178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,7168,0.011687999798191918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,1536,0.009022222624884712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,1536,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,768,0.003555555724435382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,1024,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,1024,0.007755555212497711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,5120,0.009016888837019602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,768,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,6144,0.02155911094612545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,768,0.006959111326270633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,512,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,512,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,256,0.007336888876226213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,256,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,1024,0.004233777936961916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,128,0.007330666813585494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,64,0.0028693332440323303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,128,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,32,0.002862222285734283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,512,0.003544888976547453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,32,0.005116444494989184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,2048,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,51200,0.04497155547142029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,16384,0.012532444463835822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,128,0.0031617778456873367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,51200,0.14785777197943792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,16384,0.05100889007250468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,3584,0.014821334017647637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,10240,0.011796444654464722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,256,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,8192,0.01038044442733129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,65536,0.18774222003089058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,10240,0.03242399957444932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,7168,0.009703111317422655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,6144,0.009020444419648912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,12288,0.01347644461525811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,5120,0.009385777844323052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,7168,0.024100444383091394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,4096,0.008017777568764156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,6144,0.021399999658266704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,64,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,3584,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,5120,0.01890222231547038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,4096,0.01623911162217458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,3584,0.0148044443792767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,3072,0.01354311075475481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,2048,0.005294222384691238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,8192,0.026590221458011206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,1536,0.004687110996908612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,2048,0.010782221953074137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,65536,0.05529955691761441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,768,0.003911111089918348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,1024,0.00794222205877304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,512,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,2560,0.006320888797442119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,256,0.0029173334025674393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,768,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,128,0.0028844444702068963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,512,0.006075555665625467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,2560,0.012087999946541257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,64,0.0028968888024489083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,32,0.0028977776981062363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,1024,0.004239999999602635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,1536,0.009009777671760982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,3072,0.006659555352396435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,32,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,51200,0.040700445572535195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,16384,0.01105955574247572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,65536,0.19250933329264322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,16384,0.05013244350751241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,256,0.005585777676767773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,10240,0.009080000221729279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,128,0.00498577786816491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,8192,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,65536,0.04989155464702182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,7168,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,10240,0.03247911069128249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,6144,0.011056889262464313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,51200,0.1478853358162774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,5120,0.01755199995305803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,6144,0.02164799968401591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,4096,0.00907466643386417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,5120,0.018788443671332467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,3584,0.008671110702885522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,4096,0.016200888488027785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,3072,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,8192,0.026704889204767015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,2560,0.010031110710567897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,7168,0.02417244513829549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,3072,0.013569778038395775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,2560,0.012090666426552666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,1536,0.004574222283230888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,12288,0.03807288739416335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,2048,0.01070755554570092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,1024,0.003907555507289039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,768,0.0035502223504914176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,1536,0.00927111092540953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,512,0.0032471112079090546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,3584,0.014858666393491956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,256,0.0028977776981062363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,12288,0.010085333552625444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,512,0.006043555421961679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,256,0.005473777651786804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,64,0.0028479999552170434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,128,0.00526577772365676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,32,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,64,0.0053004444473319584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,1024,0.007679999702506595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,768,0.007333333293596904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,32,0.004910222358173794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,2048,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,128,0.0028728888266616394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,12288,0.038007110357284546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,64,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,16384,2.724588394165039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,16384,0.9200400246514214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,16384,3.3742845323350696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,51200,3.4169031778971353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,16384,2.7819165123833547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,12288,2.213850657145182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,12288,0.6708746486239963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,12288,2.3919359842936196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,10240,0.5888195567660862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,12288,2.088224834865994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,10240,1.8877510494656031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,10240,2.088001675075955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,8192,1.6129866706000433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,10240,1.7367653316921656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,8192,1.4364692899915907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,8192,0.4685217539469401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,51200,9.164016723632812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,7168,1.5093332926432292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,7168,1.2441707187228732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,8192,1.3408809238009984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,6144,1.3416257434421117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,6144,0.358209768931071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,7168,1.2005991405910916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,7168,0.4177564515007867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,6144,1.0841902626885307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,6144,1.0305662155151367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,5120,0.9973555670844184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,5120,0.31210401323106557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,51200,13.126939561631945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,4096,0.6692328982883029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,4096,0.2583182122972277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,5120,0.9010978274875217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,5120,0.8374755647447375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,4096,0.7243199878268771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,3584,0.6034435696072048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,3584,0.23018132315741646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,3072,0.48875824610392254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,3584,0.6353377766079372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,4096,0.6649102105034722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,3584,0.5721199777391222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,2560,0.4133324358198378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,3072,0.4987902111477322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,2560,0.17789688375261095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,2560,0.45049333572387695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,2048,0.3530400064256456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,51200,9.759767320421007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,2560,0.42996711201137966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,2048,0.15061867237091064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,1536,0.29285865359836155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,2048,0.34873244497511124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,3072,0.20410933759477404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,1536,0.12121955553690593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,3072,0.5312222374810113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,1536,0.2804373370276557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,1024,0.09520355860392253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,1024,0.22140267160203722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,1024,0.20397510793473986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,768,0.08390755785836114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,768,0.17722400029500326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,2048,0.3843822214338515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,1024,0.21004445023006865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,512,0.13666489389207628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,256,0.08663111262851292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,1536,0.317559109793769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,768,0.14173511664072672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,128,0.0632622241973877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,256,0.06831910875108507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,256,0.1359244452582465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,512,0.11394222577412923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,64,0.061103112167782254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,128,0.06525688701205783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,128,0.12809333536359999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,32,0.05728622277577718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,512,0.15263022316826716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,64,0.1304239961836073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,256,0.10861244466569688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,768,0.16850222481621635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,128,0.08401866753896077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,512,0.07593155569500394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,32,0.13127288553449842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,51200,2.6924701266818576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,65536,3.307602564493815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,51200,6.9411731296115455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,16384,2.7490728166368275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,16384,2.7232960595024953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,65536,9.105869717068142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,16384,0.7210159831576877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,12288,1.7934710184733074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,12288,0.5379618008931478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,12288,1.8209830390082464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,16384,2.1479670206705728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,51200,10.974163479275175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,12288,1.6274933285183375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,10240,0.4382675488789876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,65536,12.972255282931856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,51200,7.5222727457682295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,10240,1.6629706488715277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,10240,1.4486604266696508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,8192,1.1117164823744032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,8192,0.39023555649651426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,10240,1.3221644295586479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,7168,1.0971315171983507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,7168,0.3215351104736328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,7168,0.9748399522569445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,8192,1.075470182630751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,6144,0.2925813462999132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,65536,9.976441277398004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,7168,0.9105901718139648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,8192,1.2193386289808485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,5120,0.7143413225809733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,5120,0.24981599383884004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,4096,0.5552808973524306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,5120,0.701932430267334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,6144,0.9143155415852865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,4096,0.6329813533359104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,4096,0.20669866932762992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,6144,0.9070630603366427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,5120,0.6266675525241429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,4096,0.5048835542466905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,3584,0.18596266375647652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,3584,0.48737064997355145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,3072,0.4102142122056749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,3584,0.458787547217475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,3072,0.16178754965464273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,2560,0.34718044598897296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,3072,0.421999110115899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,3072,0.40462401178148055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,3584,0.4345253308614095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,2560,0.14650844203101263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,6144,0.7523502243889703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,2048,0.11864711178673638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,2048,0.3047146797180176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,2048,0.2729777759975857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,1024,0.15339556005265978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,1536,0.09847999943627252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,1536,0.22153955035739473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,2048,0.2767884466383192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,768,0.12170399559868707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,1024,0.07722044653362699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,1024,0.16386755307515463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,1536,0.20285244782765707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,768,0.06816177897983126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,512,0.09283377726872762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,768,0.14005066288842097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,768,0.13345511754353842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,1536,0.2335137791103787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,512,0.060871111022101514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,512,0.12086933188968235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,2560,0.35816089312235516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,512,0.10994578070110744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,128,0.051575111018286816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,256,0.05522577630148994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,1024,0.17906399567921957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,256,0.08671377764807807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,64,0.04470488760206434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,128,0.052798221508661904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,32,0.04545155498716566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,128,0.06772622134950426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,128,0.10143555535210504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,64,0.10319911109076606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,32,0.10380266772376166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,256,0.10764088895585801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,2560,0.3349608845180935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,256,0.058676441510518394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,65536,1.0593235227796767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,65536,2.8308756086561413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,16384,0.8245804574754504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,51200,0.8096533351474338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,16384,0.26793866687350804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,16384,0.7929635577731662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,51200,2.423233879937066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,12288,0.542327086130778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,51200,3.2643689049614797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,16384,0.6665786637200249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,12288,0.1826195584403144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,51200,2.3576338026258683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,10240,0.4564613236321344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,12288,0.557204458448622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,12288,0.5005155669318305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,10240,0.4860951105753581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,10240,0.16841867234971789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,8192,0.3679706785413954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,65536,3.968829260932075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,8192,0.12526222070058188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,7168,0.28803910149468315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,8192,0.3683973418341742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,8192,0.3354746765560574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,7168,0.45404619640774196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,7168,0.29445332951015896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,65536,3.0319254133436413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,6144,0.27801866001553005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,6144,0.09761777851316665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,5120,0.2276373439364963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,5120,0.08440799845589532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,4096,0.1801688936021593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,5120,0.20926222536298963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,7168,0.11179733276367188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,4096,0.07000444332758586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,4096,0.18796800242529976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,6144,0.2371448940700955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,4096,0.17479911115434435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,3072,0.1407244470384386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,3584,0.16264000203874376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,3584,0.152145783106486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,5120,0.23238754272460938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,10240,0.4133306609259711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,3072,0.14236888620588514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,3072,0.05635911226272583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,3072,0.13609955045912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,2560,0.04981155527962578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,2560,0.12252888414594863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,3584,0.16112089157104492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,2560,0.11556000179714626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,3584,0.06325866778691609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,2048,0.10116977824105157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,1536,0.06978222396638659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,2048,0.0967751079135471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,2560,0.10486933257844712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,1024,0.051869332790374756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,2048,0.09457155730989243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,1024,0.062255111005571157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,2048,0.0431600014368693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,1024,0.030208001534144085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,1024,0.06264622343911065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,1536,0.035699556271235146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,768,0.05024088753594292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,1536,0.07846666706932916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,512,0.030615998639000788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,768,0.05177066723505656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,512,0.025461332665549383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,256,0.022153778208626643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,512,0.04279377725389269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,6144,0.25677066379123265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,128,0.01907111042075687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,256,0.022981334063741896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,256,0.038359112209743924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,1536,0.08097333378261991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,256,0.03362577822473314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,64,0.01633155511485206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,128,0.03621777892112732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,768,0.027752000424597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,128,0.028129776318868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,64,0.036673777633243136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,512,0.04292710953288608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,32,0.0366924438211653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,768,0.03839110996988084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,128,0.02219644520017836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,65536,0.7694791158040365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,65536,2.0717254214816623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,51200,1.8332862854003906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,32,0.016200888488027785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,16384,0.5127653545803493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,51200,0.6822880109151205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,65536,2.977064980400933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,51200,2.423771540323893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,12288,0.4309973451826308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,16384,0.6402026812235514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,16384,0.4910160170661078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,12288,0.14003911283281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,10240,0.37958134545220273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,12288,0.42878132396274143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,10240,0.36823288599650067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,12288,0.3754124376508925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,8192,0.2437031004163954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,10240,0.30968356132507324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,8192,0.09683466619915432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,7168,0.2460284498002794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,8192,0.25419643190171987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,65536,2.266802681816949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,7168,0.2472942140367296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,7168,0.08833155367109512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,6144,0.21451466613345674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,7168,0.22575200928582084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,5120,0.17677599853939482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,6144,0.21312533484564888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,6144,0.1944355567296346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,8192,0.2839164468977186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,5120,0.06695022185643514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,5120,0.17889511585235596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,51200,1.7501458062065973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,5120,0.16200889481438532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,10240,0.12336533599429661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,4096,0.05466399921311272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,4096,0.1326462162865533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,16384,0.1805760065714518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,3584,0.04925866590605842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,3584,0.12552621629503039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,3584,0.1212586694293552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,4096,0.13911733362409803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,3072,0.11019644472334121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,2560,0.09221155775917901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,3072,0.10586844550238715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,4096,0.14201333787706164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,2560,0.03943822118971083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,2560,0.09488622347513835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,6144,0.07749511135949029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,3072,0.10080533557467991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,2048,0.03369244601991441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,1536,0.056158224741617836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,2048,0.07642133368386163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,1536,0.02735111117362976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,1024,0.03633244501219855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,1536,0.06252444452709623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,3584,0.12451822227901882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,2048,0.07369244760937162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,2560,0.09142666392856175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,768,0.028423110644022625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,1024,0.04861244559288025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,2048,0.078456891907586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,768,0.021344888541433547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,768,0.038154668278164335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,768,0.04040622380044725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,3072,0.04399822155634562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,512,0.031494223409228854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,256,0.017176000608338248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,512,0.018882667024930317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,512,0.03420799970626831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,1024,0.04749155706829495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,256,0.017810665898852877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,256,0.028344889481862385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,256,0.027254222167862788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,512,0.022638221581776936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,64,0.013253333667914072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,128,0.027144889036814373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,128,0.017834666702482436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,128,0.02327288852797614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,1536,0.06270311276117961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,64,0.02665511104795668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,32,0.02682311170630985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,1024,0.02271822260485755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,128,0.015425778097576566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,32,0.01312800000111262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,65536,0.6404275364345974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,65536,1.8470177120632594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,51200,1.4655413097805445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,51200,0.5014328956604004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,16384,0.4663031366136339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,16384,0.49303732977973086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,65536,2.626984066433377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,51200,2.12099732293023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,16384,0.415829340616862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,12288,0.12380799982282852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,65536,1.952007081773546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,10240,0.25433778762817383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,12288,0.3175857861836751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,51200,1.4712675942314997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,10240,0.10103466775682236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,8192,0.21432888507843018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,10240,0.29951466454399955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,10240,0.2699466546376546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,12288,0.3634755611419678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,8192,0.30490043428209096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,8192,0.21076977252960205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,12288,0.3603297869364421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,7168,0.20976532830132377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,7168,0.18550666173299155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,6144,0.06458489100138347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,16384,0.15326310528649226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,6144,0.16660711500379774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,5120,0.15425599945916071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,8192,0.08332355817159016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,5120,0.0557875567012363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,5120,0.15079555246565077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,6144,0.15053333176506892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,4096,0.12188000149197048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,4096,0.04654399885071648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,3584,0.10248888863457574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,6144,0.1817795568042331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,3584,0.04155288802252875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,3584,0.1070293320549859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,3072,0.09061600102318658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,3584,0.10189333226945664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,3072,0.036963555547926165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,3072,0.09386577871110703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,2560,0.07691199911965264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,3072,0.09088889095518325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,4096,0.116321775648329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,2560,0.03359377715322707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,2048,0.061266667313045926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,2560,0.07823644744025336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,7168,0.2102968825234307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,1536,0.04712355468008253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,2048,0.028761777612898085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,2048,0.06755733489990234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,5120,0.14059821764628092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,1536,0.053532444768481784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,1536,0.024159999357329473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,1536,0.05411822266048855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,1024,0.020311999652120803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,1024,0.040821333726247154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,768,0.02318222158484989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,1024,0.04300089014901055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,2048,0.06681155496173434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,512,0.018560889694425795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,768,0.035736001200146146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,7168,0.07475111219618055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,1024,0.0307262208726671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,256,0.014740443891949125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,512,0.016566221912701923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,512,0.02983022067281935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,2560,0.08130666944715711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,256,0.015980444020695157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,128,0.012076444096035428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,256,0.02370311154259576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,4096,0.12233866585625543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,128,0.023355555203225877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,128,0.02035199933581882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,32,0.011635555161370171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,64,0.023410666320059035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,32,0.023071999351183575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,256,0.024119999673631456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,768,0.01871466636657715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,128,0.015670221712854173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,512,0.027383999692069158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,65536,0.5327439837985569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,768,0.03328088919321696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,64,0.011100444528791638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,65536,1.972683588663737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,51200,1.2418284946017795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,16384,0.3232853412628174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,51200,0.42923733923170304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,65536,1.5331617991129558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,16384,0.12838755713568792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,16384,0.47988001505533856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,12288,0.27559465832180446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,12288,0.10116444693671332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,65536,1.3056862089369032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,12288,0.2984568807813856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,51200,1.166683514912923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,12288,0.2544231149885389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,10240,0.08540355496936375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,10240,0.24151200718349883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,10240,0.21255911721123588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,8192,0.19545688894059923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,7168,0.16874310705396864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,10240,0.24347376823425293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,7168,0.15256533357832167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,6144,0.1447804503970676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,8192,0.0698924462000529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,16384,0.3422791163126628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,6144,0.14885066615210638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,6144,0.05345777670542399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,7168,0.1725191142823961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,6144,0.1359271075990465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,8192,0.1921235587861803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,5120,0.04691733254326714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,4096,0.08277689086066352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,5120,0.11470044983757867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,51200,1.6729502148098414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,4096,0.03828533490498861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,4096,0.09967466857698228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,4096,0.0950088898340861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,5120,0.11934755908118354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,3584,0.03427111109097799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,3584,0.08875644207000732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,7168,0.06050399939219157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,5120,0.1236835585700141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,3072,0.07784089114930895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,2560,0.05670577949947781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,3072,0.07553244299358793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,8192,0.17637244860331217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,2560,0.06720266739527385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,2048,0.05284888876809014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,2560,0.06647644440333049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,3072,0.0748515526453654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,2048,0.02453688946035173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,3584,0.08505688773261176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,2048,0.055345777008268565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,3072,0.03133866522047255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,1536,0.020405333903100755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,1536,0.044975111881891884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,1536,0.04564355479346382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,2560,0.02823466724819607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,3584,0.08792089091406928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,768,0.019744000501102872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,1536,0.03626666797531976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,768,0.027792887555228338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,512,0.015185778339703878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,768,0.015604444675975375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,1024,0.02504088812404209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,512,0.02311555544535319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,512,0.014496889379289416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,512,0.025449777642885845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,1024,0.017536888519922893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,1024,0.03608266512552897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,128,0.010994666980372535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,256,0.013467555244763693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,256,0.020228443874253165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,2048,0.05583733320236206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,64,0.009348444640636444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,768,0.030227555168999568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,128,0.013730666703648038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,32,0.009454222189055549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,128,0.01758044461409251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,64,0.019500444332758587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,32,0.019494222270117868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,256,0.011843555503421359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,1024,0.03438933359252082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,256,0.020118221640586853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,65536,0.49885956446329754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,128,0.019654222660594516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,65536,1.458099577162001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,51200,0.9947972827487521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,16384,0.3104337851206462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,51200,0.3927440113491482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,16384,0.11364177862803142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,16384,0.367232879002889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,51200,1.353957282172309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,16384,0.3065360122256809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,65536,1.3390240139431422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,12288,0.08891555335786606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,10240,0.2367475562625461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,51200,1.0075617896185982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,12288,0.22562932968139648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,10240,0.23697956403096518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,65536,1.8449812995062933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,12288,0.26732444763183594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,8192,0.17258932855394152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,8192,0.060976889398362905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,7168,0.16693422529432508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,12288,0.2548915545145671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,7168,0.054494222005208336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,7168,0.15128355556064182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,6144,0.12632889217800564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,10240,0.2044995493359036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,6144,0.13058311409420437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,6144,0.12074044015672471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,8192,0.157733334435357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,5120,0.1102657781706916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,10240,0.08049955632951525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,4096,0.08453333377838135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,5120,0.1028675569428338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,4096,0.03477955526775784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,4096,0.08959733115302192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,6144,0.048954665660858154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,3584,0.08327822552786933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,4096,0.08660178052054511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,5120,0.11027110947502984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,3072,0.06704266866048177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,3584,0.030689779255125258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,3584,0.07970133092668322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,3584,0.07595377498202853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,3072,0.07056978013780382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,8192,0.18776978386773002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,3072,0.06853333446714613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,7168,0.1412284506691827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,2048,0.04666399955749512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,2560,0.02571822206179301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,2560,0.06030311187108358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,1536,0.03405866689152188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,2560,0.06061244673199124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,2048,0.022322666313913133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,1536,0.04028266668319702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,1536,0.018573333819707234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,1024,0.021122665868865118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,1536,0.042077332735061646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,2560,0.05409155620468987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,768,0.020204444726308186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,1024,0.01576266686121623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,5120,0.04156088829040527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,1024,0.031046221653620403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,512,0.015838222371207345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,768,0.014486221803559197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,768,0.027980443504121568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,2048,0.05032444331381056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,256,0.012768000364303589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,512,0.013748444616794586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,512,0.022740445203251306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,3072,0.027801778581407335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,256,0.013014222184816996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,128,0.010503110786279043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,256,0.019332443674405415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,1024,0.03230133321550157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,128,0.017851556340853374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,64,0.009384888741705153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,128,0.012798222402731577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,128,0.017245334055688646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,2048,0.05084800057941013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,64,0.017599110802014668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,32,0.017529777354664274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,256,0.018363555272420246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,65536,0.40828977690802676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,768,0.025462221768167283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,32,0.009668444593747457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,512,0.020675554871559143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,51200,0.7767786449856229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,16384,0.27997422218322754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,65536,1.154062271118164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,16384,0.3164373238881429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,65536,1.1981742646959093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,12288,0.1975111166636149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,51200,0.8733217981126574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,12288,0.08425155613157485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,12288,0.22250488069322374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,51200,0.3332088788350423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,12288,0.1975004408094618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,10240,0.18341422080993652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,10240,0.06615377797020806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,8192,0.1465244425667657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,10240,0.18716977702246773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,10240,0.1675644450717502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,16384,0.10777599944008721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,8192,0.0547653337319692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,7168,0.12765422132280138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,8192,0.13583021693759495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,7168,0.048199110560946994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,6144,0.10619466834598118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,65536,1.430695957607693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,7168,0.12180177370707194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,6144,0.04148355457517836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,6144,0.11501067214541966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,6144,0.10597778028912014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,16384,0.2614906628926595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,5120,0.09667733642790054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,8192,0.15092355675167507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,4096,0.06876444154315524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,5120,0.09062400129106309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,7168,0.13408888710869685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,4096,0.07862666580412123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,3584,0.0672266681989034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,4096,0.07658844523959689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,5120,0.09500088956620957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,3072,0.05366666780577766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,3584,0.027884443600972492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,3584,0.07008444600635104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,51200,1.2330693138970268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,2560,0.042066666815016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,3072,0.06167911158667671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,5120,0.0357528891828325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,3072,0.06116177638371786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,2048,0.035952889257007174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,2560,0.023020444644822016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,4096,0.03030666708946228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,2560,0.05267288949754503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,1536,0.028689778513378564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,2048,0.02025422288311852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,2048,0.04470400015513102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,3584,0.06778044170803495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,1536,0.017258667283587985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,1024,0.021288888321982488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,1536,0.037280887365341187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,1024,0.014697778556081982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,1024,0.027569777435726587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,1024,0.029267556137508813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,768,0.013796444568369122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,768,0.022816000713242426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,512,0.012296888563368054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,1536,0.03586488962173462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,512,0.018726223044925265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,256,0.009544000029563904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,512,0.012488000094890594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,768,0.016505777835845947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,2560,0.054042667150497437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,128,0.008805333740181392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,256,0.011739555332395764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,2048,0.04488889045185513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,768,0.024989333417680528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,64,0.008000888758235509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,128,0.015855999456511605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,128,0.015964445140626695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,3072,0.025017778078715008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,512,0.020984888076782227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,64,0.015855999456511605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,32,0.015841777125994366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,256,0.016571554872724745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,256,0.017631111873520743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,65536,0.8740666707356771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,65536,0.3770311143663194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,128,0.011791110866599612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,32,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,51200,0.7886204189724393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,51200,0.2972533437940809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,65536,1.2670364379882812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,65536,0.9248453776041666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,51200,0.9570870929294162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,16384,0.2548444535997179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,16384,0.24738756815592447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,12288,0.06666400035222371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,12288,0.2401182121700711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,12288,0.17350755797492134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,10240,0.15987377696567112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,10240,0.14625155925750732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,8192,0.13938844203948975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,12288,0.18492978148990205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,8192,0.04640000065167745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,51200,0.7329742113749186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,7168,0.12265777587890625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,8192,0.12054310904608832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,16384,0.2319413291083442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,7168,0.04197955462667677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,7168,0.1145697832107544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,10240,0.05866666634877523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,7168,0.10539378060234918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,16384,0.08703466918733384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,6144,0.0369120008415646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,6144,0.0990133351749844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,6144,0.09175999959309895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,4096,0.06763466861512926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,8192,0.1310746669769287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,5120,0.08425422509511311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,5120,0.0790133343802558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,4096,0.027662222584088642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,4096,0.06828977664311726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,6144,0.09772888819376628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,4096,0.06615644693374634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,3072,0.04938755432764689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,3584,0.024716445141368445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,3584,0.06091644366582235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,5120,0.09133422374725342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,3584,0.05904889106750488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,2560,0.04263733492957222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,3072,0.02239111065864563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,3072,0.05360888772540622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,10240,0.17283556196424696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,2560,0.046560890144772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,3584,0.05961333380805122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,2560,0.04625511169433594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,2048,0.03904711206754049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,2048,0.017887110511461895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,1536,0.026736888620588515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,2048,0.03947822252909342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,1536,0.031228442986806233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,1024,0.0200311111079322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,5120,0.032538665665520564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,2048,0.03504000107447306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,1024,0.024531554844644334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,768,0.01610844499535031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,1024,0.02583911187118954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,768,0.012406222522258759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,512,0.012376888758606382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,768,0.02201777696609497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,1536,0.032623999648623996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,3072,0.05339822173118591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,256,0.009729777773221334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,512,0.011756444142924415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,512,0.018276444739765592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,256,0.014694222145610385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,256,0.011425777441925473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,768,0.020073778099483915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,1536,0.015841777125994366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,2560,0.020503999458418954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,128,0.015101333459218344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,64,0.007293333609898885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,32,0.007276444799370236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,64,0.013921777407328287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,256,0.016192888220151264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,32,0.013863999810483722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,128,0.014169777433077494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,65536,0.6508240169949001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,128,0.011083555718262991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,512,0.016586666305859882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,1024,0.013432888521088494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,51200,0.5063929027981228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,128,0.008016889293988546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,65536,0.7584942181905111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,51200,0.7640835444132487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,16384,0.21077956093682182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,16384,0.075436446401808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,16384,0.19663111368815103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,65536,0.3040613333384196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,65536,1.025718265109592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,12288,0.05633866786956787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,12288,0.1578764385647244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,10240,0.11874755223592122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,51200,0.24725421269734701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,8192,0.1033688916100396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,10240,0.048884444766574435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,10240,0.1331928835974799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,10240,0.12211022112104629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,12288,0.1288248830371433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,7168,0.08686400122112697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,8192,0.10753777954313491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,8192,0.09974133306079441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,6144,0.07127733363045587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,7168,0.034908443689346313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,7168,0.09573511282602946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,7168,0.08909510903888279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,16384,0.18618667125701904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,5120,0.0610062215063307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,6144,0.08335466517342462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,6144,0.07765422264734904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,51200,0.6019324196709527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,5120,0.07043200069003634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,5120,0.06751377715004815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,4096,0.0575431121720208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,4096,0.02348711093266805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,3584,0.041533334387673274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,4096,0.05647644731733534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,12288,0.14330222871568468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,6144,0.031614220804638334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,3584,0.0514008899529775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,3584,0.05078755484686958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,4096,0.04766311248143514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,3072,0.01994755533006456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,3072,0.045922666788101196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,8192,0.039184000757005476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,3072,0.04518577787611219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,2048,0.025359110699759588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,2560,0.018138666947682697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,2560,0.039447112215889826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,2560,0.04003466831313239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,1536,0.020633776982625324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,2048,0.01590399940808614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,3584,0.021764443980322942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,2048,0.03458133339881897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,1024,0.015397333436542086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,1536,0.014057776994175382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,5120,0.027666666441493567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,1536,0.028207110034094915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,768,0.012157333393891653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,1024,0.012151999606026543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,1024,0.021369778447681006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,1024,0.02306844459639655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,768,0.0173973325226042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,768,0.011401777466138205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,2048,0.03341599967744615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,768,0.0192266669538286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,3072,0.03694577680693732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,256,0.007584888901975419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,512,0.014495111174053617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,512,0.01072533345884747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,1536,0.02737155556678772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,128,0.006610666712125142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,256,0.010070222119490305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,256,0.012813333008024426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,256,0.014639111028777229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,64,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,128,0.013470222552617392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,32,0.006274666637182236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,64,0.012253333297040729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,32,0.012096000214417776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,512,0.01622488929165734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,65536,0.5985377629597982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,128,0.010223111344708337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,512,0.009365333451165093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,65536,0.2864169014824761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,51200,0.5331981976826986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,65536,0.8605173428853353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,65536,0.7708142068650988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,51200,0.6930418014526367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,128,0.012446222205956778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,51200,0.571307553185357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,12288,0.13197244538201228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,12288,0.150328000386556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,12288,0.1402746703889635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,2560,0.0314026673634847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,12288,0.0544124444325765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,16384,0.18953422705332437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,10240,0.12007644441392686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,16384,0.19341688685946992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,10240,0.11904711193508571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,8192,0.09740622176064385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,8192,0.09823911057578193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,10240,0.11426756117078994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,8192,0.03728444377581278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,7168,0.08647733264499241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,8192,0.0955573320388794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,6144,0.07104355759090848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,16384,0.18856355879041883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,6144,0.07459733221266004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,7168,0.08230221933788724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,5120,0.057743999693128795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,6144,0.029932445949978296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,6144,0.0767493314213223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,16384,0.07048977745903863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,7168,0.08728266424602932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,4096,0.04526666800181071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,5120,0.026704000102149114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,7168,0.03385422295994229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,5120,0.065975997183058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,3584,0.042711112234327525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,4096,0.02288533416059282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,4096,0.05196533269352383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,51200,0.22761244244045684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,3584,0.02067377832200792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,3584,0.04667199982537163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,3584,0.04991822110282051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,2560,0.030271109607484605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,3072,0.018949333164427016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,3072,0.04435911112361484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,10240,0.04750488864050972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,2048,0.02520711057715946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,5120,0.06384711133109199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,2560,0.017504889104101393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,2560,0.03851022322972616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,1536,0.020076443751653034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,2048,0.030765334765116375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,4096,0.05486399928728739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,1536,0.0244195560614268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,3072,0.0413955549399058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,1024,0.014759999182489185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,2560,0.03631822268168131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,1024,0.018805333309703402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,768,0.012479110724396177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,1024,0.01200977795653873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,3072,0.034500443273120456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,768,0.01603200038274129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,768,0.01105955574247572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,768,0.018797333041826885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,1536,0.013588444226317935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,512,0.013440888788965015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,256,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,1024,0.022125333547592163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,2048,0.015533333023389181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,256,0.01185955521133211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,2048,0.033542222446865506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,256,0.010048888623714447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,256,0.013193777865833707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,128,0.011424000064531961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,128,0.010077333284748925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,512,0.010399999717871347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,512,0.016125332978036668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,32,0.006047111004590988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,64,0.01145155562294854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,32,0.01144977741771274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,512,0.010031110710567897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,65536,0.58625242445204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,1536,0.027512888113657635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,128,0.013468444347381592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,65536,0.26467821333143443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,51200,0.45043288336859805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,65536,0.7469395531548394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,16384,0.14335021707746717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,51200,0.6202800008985732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,64,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,16384,0.1687715583377414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,12288,0.09513688749737209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,16384,0.06672533353169759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,12288,0.05163377854559156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,12288,0.12616977426740858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,10240,0.1014782190322876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,51200,0.21876355012257895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,10240,0.04636977778540718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,10240,0.10718488693237305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,51200,0.48327912224663633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,10240,0.0986444420284695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,16384,0.1515457762612237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,7168,0.06808000140719943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,8192,0.08675644132826064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,8192,0.08089955647786458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,65536,0.5908195707533095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,12288,0.11560000313652886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,7168,0.07331555419498019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,8192,0.08164711130989923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,6144,0.0648017790582445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,6144,0.029515554507573444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,8192,0.0353999998834398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,128,0.006650666809744305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,5120,0.0571955574883355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,7168,0.07788622379302979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,4096,0.03917155663172404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,5120,0.025267556309700012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,4096,0.02198666665289137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,4096,0.046685334708955556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,3584,0.03287377622392442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,4096,0.047673778401480786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,5120,0.04573333263397217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,3072,0.028288000159793433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,3584,0.041837334632873535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,3584,0.04194666610823738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,7168,0.03243733445803324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,3072,0.03755022088686625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,2560,0.023392000132136877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,3072,0.03760533200369941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,6144,0.06688621971342298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,2560,0.03251377741495768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,2560,0.03271466824743483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,3584,0.020218667056825426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,2048,0.027253333065244887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,1536,0.01587199999226464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,2048,0.028455999162462022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,6144,0.05535644292831421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,3072,0.018391999933454726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,1536,0.02224622170130412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,1536,0.012440000143316058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,1536,0.023404444257418316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,2560,0.016536888149049547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,2048,0.020020445187886555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,1024,0.011157333023018308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,1024,0.017198221551047433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,1024,0.018911111685964797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,512,0.007671111159854465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,768,0.01071466671095954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,768,0.0162426663769616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,5120,0.05515911181767782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,512,0.012192000117566852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,512,0.010027555955780877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,512,0.014126222994592456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,1024,0.012048888537618848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,128,0.005985777825117111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,256,0.009364444348547194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,256,0.010829333629873065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,256,0.01314933349688848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,64,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,128,0.012415111064910889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,32,0.005568888866239124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,2048,0.014322666658295525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,64,0.010416888528399997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,256,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,32,0.010435555544164447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,65536,0.5046844482421875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,768,0.009899555808968013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,65536,0.22943467564053008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,128,0.010614222122563256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,65536,0.6467031372918023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,51200,0.40398399035135907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,16384,0.13316978348626032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,65536,0.580741352505154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,51200,0.5211706691318089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,16384,0.16771377457512748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,51200,0.4540088971455892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,16384,0.14924444092644587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,12288,0.1134986612531874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,768,0.014831110835075378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,12288,0.11344622241126166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,12288,0.04533955454826355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,10240,0.08353066444396973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,51200,0.19077866607242158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,8192,0.07283733288447063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,16384,0.057914667659335665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,10240,0.09883999824523926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,8192,0.07659022013346355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,8192,0.0311173333062066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,128,0.009512888888518015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,7168,0.06789688931571113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,7168,0.07150222195519342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,10240,0.039686222871144615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,7168,0.028967999749713477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,10240,0.0936524470647176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,6144,0.025153777665562097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,6144,0.05881600247489082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,5120,0.04276266694068909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,6144,0.06452355782190959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,5120,0.022779555784331426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,5120,0.05008088880115085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,4096,0.033803555700514056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,5120,0.05428266525268555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,3584,0.030305776331159804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,4096,0.01937599976857503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,6144,0.04888800117704603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,4096,0.0418542226155599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,4096,0.04562222295337253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,3072,0.02715822226471371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,3584,0.017779555585649278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,3584,0.037125332487954035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,2560,0.022837332553333704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,3072,0.016572443975342643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,7168,0.06102933486302694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,3072,0.03762311074468825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,2048,0.019518221418062847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,2560,0.031814220878813006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,12288,0.09881244765387641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,2048,0.024336000283559162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,1536,0.015847111741701763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,2048,0.028010666370391846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,3584,0.041293332974116005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,8192,0.07971200015809801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,1536,0.019854222734769184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,1536,0.023185777995321486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,2560,0.028669334120220605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,1024,0.015522667103343539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,1024,0.010405333505736457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,768,0.010008888940016428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,1024,0.01851111153761546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,2048,0.013919111755159167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,768,0.01349955548842748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,768,0.0162764439980189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,3072,0.0332551101843516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,1024,0.011920889218648275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,512,0.011111111276679568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,512,0.009476444787449306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,512,0.013752000199423896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,256,0.010075555907355415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,128,0.005382222019963794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,768,0.010027555955780877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,512,0.009003555609120263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,128,0.009706666900051964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,64,0.004959110998445087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,1536,0.012075555821259817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,32,0.005285333428117964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,64,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,32,0.009714666340086196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,256,0.012216000093354119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,65536,0.39347110854254824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,128,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,256,0.007303111255168915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,65536,0.4383217758602566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,51200,0.2970213360256619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,256,0.00905688852071762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,51200,0.1375759972466363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,16384,0.09505955378214519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,51200,0.4120373460981581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,51200,0.34387734201219344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,16384,0.04561511013242933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,65536,0.17398399776882598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,16384,0.11650044388241237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,128,0.012320888539155325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,10240,0.06249777475992838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,12288,0.088392886850569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,10240,0.08094488912158541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,10240,0.07578488853242663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,12288,0.07928978072272407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,8192,0.05016977919472588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,2560,0.014859555496109856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,8192,0.02442666722668542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,12288,0.03570311268170675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,12288,0.09615821970833673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,8192,0.06314400169584486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,7168,0.022308443983395893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,7168,0.05857333209779528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,7168,0.05602755811479357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,16384,0.13088888592190212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,65536,0.5132826699150933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,6144,0.05067200130886502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,5120,0.03096888793839349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,6144,0.0203146669599745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,8192,0.06605155600441827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,5120,0.01794488893614875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,5120,0.042968001630571157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,4096,0.02591288917594486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,5120,0.04358844293488396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,6144,0.03850311040878296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,3584,0.022643556197484333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,4096,0.01571377780702379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,10240,0.03096710973315769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,4096,0.03582133187188043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,4096,0.03629688753022088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,3584,0.014484445254007975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,3584,0.032129777802361384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,3584,0.033415112230512835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,7168,0.046760890218946666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,2560,0.01756622228357527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,3072,0.0299653344684177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,2560,0.01242844429281023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,2048,0.014961777461899651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,2560,0.02443555494149526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,2560,0.02630222174856398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,1536,0.01218933363755544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,2048,0.02118844456142849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,2048,0.02216622233390808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,1536,0.017239999439981248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,1024,0.00962933318482505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,6144,0.05008977651596069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,3072,0.013342222405804528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,1024,0.009400000174840292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,1024,0.013487999637921652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,1024,0.015527110960748462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,2048,0.011376889215575324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,768,0.011755555868148804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,3072,0.02022133270899455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,768,0.009015999734401703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,768,0.01367288910680347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,1536,0.01038577738735411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,256,0.005405333307054307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,3072,0.028579556279712256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,512,0.00998399986161126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,512,0.011892444557613797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,128,0.004946666873163647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,256,0.008370666868156856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,256,0.009022222624884712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,256,0.011062222222487131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,64,0.004222222086456087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,128,0.008353778057628209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,32,0.004563555535342959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,128,0.010746666954623329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,64,0.00832088871134652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,32,0.0086986662613021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,65536,0.3083004421657986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,768,0.008018666671382057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,65536,0.4148417843712701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,128,0.008340444829728868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,65536,0.42362221082051593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,512,0.006933333145247565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,16384,0.08018666505813599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,51200,0.3265635437435574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,512,0.00867377801073922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,51200,0.3331955538855659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,16384,0.10723555750317043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,12288,0.05949866771697998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,65536,0.1683057811525133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,16384,0.1127199994193183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,10240,0.04822311136457655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,12288,0.03496355480617947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,12288,0.07981599701775445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,12288,0.08681866857740615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,10240,0.0306915541489919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,51200,0.12792444229125977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,1536,0.018616888258192275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,10240,0.07586222224765353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,7168,0.0328746669822269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,16384,0.0443466669983334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,8192,0.0614524417453342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,7168,0.048419555028279625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,51200,0.25541867150200737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,7168,0.05476266807980008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,8192,0.03979911075698005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,6144,0.04235022266705831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,5120,0.024193776978386775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,8192,0.024271999796231587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,5120,0.03626399901178148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,7168,0.021761778328153823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,5120,0.042135109504063926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,4096,0.020460445019933913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,10240,0.06719822353786893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,4096,0.030456887351142034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,4096,0.015227556228637695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,4096,0.035641776190863714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,8192,0.05443911088837517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,3584,0.027200889256265428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,3072,0.01571111049917009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,6144,0.028014222780863445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,3072,0.02401777770784166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,6144,0.048713776800367564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,3072,0.028471999698215063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,2560,0.013665777941544851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,3584,0.014338667194048563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,3584,0.03219644559754266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,5120,0.017316444052590262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,2560,0.02118844456142849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,2560,0.012142221960756512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,2560,0.02553333342075348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,2048,0.01806400054030948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,6144,0.01959022217326694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,3072,0.01334755536582735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,1536,0.014929778046078153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,1536,0.010055999788973067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,1536,0.0184906671444575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,1024,0.007954667011896769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,1024,0.012091555529170565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,3584,0.01825066738658481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,768,0.006830222076839871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,768,0.01039644413524204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,2048,0.01166044423977534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,768,0.009069333473841349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,768,0.013602666556835175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,2048,0.01107022249036365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,512,0.008989333278603023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,256,0.004597333156400257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,1024,0.009375111096435124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,512,0.01182933317290412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,256,0.00832444429397583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,128,0.004313777718279096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,256,0.01054133309258355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,512,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,128,0.0073520003093613525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,128,0.010442666709423065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,1536,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,32,0.0042462220622433555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,512,0.008438222110271454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,64,0.007618666523032718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,65536,0.23536178800794813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,256,0.008016000191370646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,65536,0.11614667044745551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,65536,0.3274159961276584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,128,0.008113777471913232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,64,0.0042257776690853965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,2048,0.021703110800849065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,16384,0.05425777700212267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,51200,0.09251555469301011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,51200,0.25983111063639325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,51200,0.22056178251902261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,32,0.007675555845101674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,12288,0.04361688759591845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,16384,0.07833421892589994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,12288,0.06336889002058241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,51200,0.1527413394716051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,10240,0.03296533226966858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,65536,0.28234489758809406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,1024,0.01516177753607432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,8192,0.0274311105410258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,10240,0.02259911100069682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,10240,0.05345866746372647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,10240,0.05093422200944689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,7168,0.022423999177085027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,8192,0.017535110314687092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,16384,0.03197777933544583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,8192,0.04302044378386604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,6144,0.01997777819633484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,7168,0.015984889533784654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,7168,0.03905244337187873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,7168,0.03846844368510776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,6144,0.014835556348164877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,6144,0.03422222203678555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,6144,0.034373333056767784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,5120,0.029206222958034937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,5120,0.013473778135246701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,4096,0.014526221487257214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,5120,0.029891557163662378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,16384,0.0841475526491801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,3584,0.01310488912794325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,4096,0.01219911128282547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,4096,0.024383111132515803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,4096,0.02552088929547204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,3072,0.011854222251309289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,3584,0.01145333300034205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,5120,0.01722400055991279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,3584,0.02276444435119629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,2560,0.010377777947319878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,3072,0.019849777221679688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,3072,0.020641777250501845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,8192,0.04379111197259691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,2560,0.017331555485725403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,2048,0.008681777450773451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,2560,0.01867822143766615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,2048,0.009675555759006077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,12288,0.025516443782382544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,1536,0.007619555625650618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,3072,0.01075111081202825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,1536,0.012444444828563266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,1536,0.014148443937301636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,1536,0.008999110923873054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,2560,0.01033688916100396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,1024,0.01033333357837465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,2048,0.015055111712879606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,12288,0.059602664576636426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,768,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,1024,0.008383110993438298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,1024,0.01201777739657296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,512,0.004910222358173794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,768,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,768,0.00906044410334693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,1024,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,768,0.01112711098459032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,512,0.007689778175618913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,512,0.010372444159454769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,256,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,256,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,256,0.00905600024594201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,3584,0.022030222747060988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,128,0.007358222372002072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,64,0.003546666767862108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,128,0.00904888908068339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,32,0.0036159998012913596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,256,0.004291555533806483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,512,0.007904888855086433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,64,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,65536,0.17967022789849174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,32,0.00700444479783376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,51200,0.13969600200653076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,128,0.0038622220357259116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,65536,0.1129831075668335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,2048,0.016487111647923786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,16384,0.046259555551740855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,51200,0.09196177456114028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,128,0.00665155549844106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,51200,0.22026666005452475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,12288,0.032151109642452665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,51200,0.21828533543480766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,16384,0.07638221979141235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,12288,0.025197333759731714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,10240,0.025480889611774023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,12288,0.05829066700405545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,10240,0.022424888279702928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,10240,0.04665333363744947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,8192,0.024352888266245525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,65536,0.28511643409729004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,65536,0.2747182316250271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,7168,0.018830221560266282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,8192,0.017290666699409485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,8192,0.03830044468243917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,16384,0.030797332525253296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,8192,0.0422017772992452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,6144,0.016030222177505493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,7168,0.016010666886965435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,7168,0.03455200129085117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,5120,0.014169777433077494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,6144,0.014656888114081489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,6144,0.03036266565322876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,5120,0.026105778084860906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,4096,0.013376889129479727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,5120,0.029480000336964924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,16384,0.07302221986982557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,4096,0.021596444977654353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,3584,0.010712888505723743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,4096,0.012093333734406365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,12288,0.0549813343418969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,4096,0.025008888708220586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,3072,0.010769777827792697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,3584,0.01146577795346578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,6144,0.03317155440648397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,3584,0.019650666250122916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,2560,0.008568888737095727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,10240,0.05059022373623318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,3072,0.01771288944615258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,3072,0.020631111330456205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,2048,0.007670222057236566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,2560,0.010108444425794814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,2560,0.018383999665578205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,2048,0.00962311112218433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,7168,0.037738667594061956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,2048,0.016362667083740234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,3584,0.022572444544898138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,1536,0.011425777441925473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,1024,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,1536,0.013839999834696451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,2560,0.015569777952300178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,1024,0.00942844483587477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,1024,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,768,0.004898666507667965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,1024,0.011995555626021491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,2048,0.013609777722093793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,1536,0.006787555499209299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,768,0.008040000167157915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,768,0.008340444829728868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,768,0.01073333372672399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,256,0.003879111260175705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,1536,0.00905600024594201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,512,0.010055999788973067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,256,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,256,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,256,0.00870577742656072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,512,0.00424799985355801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,128,0.006231110956933763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,128,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,5120,0.013455111119482251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,128,0.009178666604889764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,32,0.0035742223262786865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,64,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,128,0.003702222059170405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,32,0.006333333336644703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,65536,0.13261332776811388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,65536,0.11175466908348931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,65536,0.23897155125935873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,51200,0.1046302252345615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,512,0.007677333222495184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,16384,0.03347911106215583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,51200,0.18936444653405082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,51200,0.20726577440897623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,16384,0.061347557438744434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,12288,0.025793777571784124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,16384,0.07232355409198336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,512,0.007335999773608313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,12288,0.04687822196218702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,10240,0.019704888264338177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,12288,0.055903110239240855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,12288,0.023738667368888855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,51200,0.08932533529069688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,8192,0.015415999624464246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,10240,0.021365332934591506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,10240,0.03979555434650845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,10240,0.047994666629367404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,16384,0.030072887738545735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,7168,0.013829333086808523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,8192,0.01664977769056956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,8192,0.04030044542418586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,7168,0.015328889091809591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,7168,0.029878222280078467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,7168,0.03604177633921305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,6144,0.026291555828518335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,6144,0.014144889182514615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,6144,0.0317884451813168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,5120,0.02238666680124071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,5120,0.027489778068330552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,5120,0.012868444124857584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,6144,0.012441777520709567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,3072,0.01073333372672399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,4096,0.01165066659450531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,3584,0.00888533310757743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,5120,0.011069333387745751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,3584,0.017197334104114108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,3584,0.02165866725974613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,3072,0.007977777885066139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,3584,0.011076444553004371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,4096,0.009355555805895064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,65536,0.26337512334187824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,3072,0.010392888552612728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,3072,0.015627556376987033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,4096,0.018887110882335238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,4096,0.023694222172101338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,2048,0.006305777778228124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,2560,0.013824889229403602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,2560,0.017675555414623685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,8192,0.03318311108483209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,2048,0.012148444851239523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,2048,0.015497777197096082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,64,0.0035333335399627686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,1536,0.010407999985747868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,2560,0.007284444239404466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,1024,0.004597333156400257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,1536,0.00869955536392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,3072,0.019641778535313077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,768,0.004230222354332606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,1024,0.00833777752187517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,1024,0.008640888664457533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,1024,0.01145066652033064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,512,0.003927111211750242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,768,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,2048,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,768,0.01054933336046007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,256,0.0035991109907627106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,512,0.006695999867386288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,512,0.009753777749008602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,1536,0.013474666410022311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,256,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,128,0.0031991110493739447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,256,0.008678221868144141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,2560,0.01001422190003925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,64,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,128,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,128,0.005664888769388199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,32,0.003237333355678452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,128,0.008886222210195329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,1536,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,65536,0.07650844256083171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,64,0.005611555443869696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,512,0.0074106670088238185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,32,0.00565155554148886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,65536,0.11088889174991184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,65536,0.19013599554697672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,16384,0.025395555628670588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,65536,0.25759734047783744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,51200,0.1485511064529419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,51200,0.20345155398050943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,16384,0.05066755414009094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,768,0.007636444436179266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,12288,0.018065777089860704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,16384,0.07079111205206977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,12288,0.023045332895384893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,12288,0.03889333208401998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,12288,0.05459022190835741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,256,0.005991999887757831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,51200,0.08898044294781154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,10240,0.03288888931274414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,10240,0.02200088898340861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,10240,0.04654933346642388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,16384,0.029527111185921565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,8192,0.02720444401105245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,8192,0.03943644298447503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,51200,0.06222044759326511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,10240,0.016220443778567843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,6144,0.01074044406414032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,7168,0.015184000134468079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,7168,0.03499111202028062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,8192,0.013915555344687568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,5120,0.00977955593003167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,6144,0.014173333843549093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,6144,0.030750221676296655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,8192,0.016551999582184684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,5120,0.012784889174832238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,5120,0.019181332654423185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,5120,0.02699911097685496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,3584,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,4096,0.01144177797767851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,4096,0.0162435554795795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,4096,0.023019555542204116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,3072,0.0070168889231152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,3584,0.011071999867757162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,3584,0.01499022212293413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,6144,0.02200977835390303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,3584,0.021020443903075323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,3072,0.010452444354693094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,7168,0.012355555262830524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,4096,0.008362666600280339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,2560,0.012095111111799875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,2048,0.005548444473081165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,2560,0.010092444717884064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,2560,0.01721333298418257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,7168,0.02440888848569658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,1536,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,2048,0.010781333678298526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,2048,0.015306666493415833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,1024,0.004228444563017951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,1536,0.008678221868144141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,2560,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,3072,0.01383733335468504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,768,0.003933333274390963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,3072,0.018929777873886954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,1024,0.01108000013563368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,768,0.0069982219073507525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,768,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,2048,0.009416888985368941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,1536,0.009053332938088311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,512,0.00600355573826366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,256,0.003535110917356279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,512,0.00740799970097012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,1536,0.013157332936922709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,256,0.005586666779385672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,128,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,256,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,512,0.003593777616818746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,768,0.01015377789735794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,64,0.003192000091075897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,128,0.0070186663005087115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,32,0.003216888962520493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,512,0.009727111293209923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,65536,0.057632002565595836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,32,0.0052800000541739995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,256,0.007275555696752336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,65536,0.18807733058929443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,65536,0.10998221900728013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,51200,0.046887109676996865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,65536,0.25643377833896214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,128,0.005193777796294955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,128,0.008605333666006723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,51200,0.088919997215271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,64,0.0052800000541739995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,51200,0.14813600646124944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,1024,0.008024888734022776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,16384,0.051148444414138794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,16384,0.07026666402816772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,12288,0.037907554043663874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,10240,0.013056889176368713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,12288,0.022808889547983806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,16384,0.01423111061255137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,10240,0.0326204432381524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,8192,0.012827555338541666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,10240,0.04616533385382759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,1024,0.007866666548781926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,16384,0.028784887658225164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,7168,0.011633777783976661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,8192,0.038562665383021034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,12288,0.05419377817047966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,7168,0.02404888967672984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,7168,0.015207111835479736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,6144,0.010377777947319878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,10240,0.020917332834667627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,12288,0.014124444789356656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,8192,0.016527111331621807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,5120,0.00923466682434082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,6144,0.030284444491068523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,5120,0.018912000788582694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,5120,0.026480888326962788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,4096,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,7168,0.03459111187193129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,51200,0.2020773357815213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,3584,0.007288888924651676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,4096,0.016177778442700703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,8192,0.02700000007947286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,4096,0.022709333234363135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,3072,0.006597333484225803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,3584,0.020688888099458482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,3072,0.013440888788965015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,5120,0.012800888882742988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,2560,0.00620177760720253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,3072,0.010390222072601318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,4096,0.011447110937701331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,6144,0.014002665877342224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,2560,0.012101333174440594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,2048,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,2560,0.009768889182143742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,3584,0.01089511149459415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,1536,0.004867555366622077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,2048,0.009343110852771336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,2048,0.010712888505723743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,2048,0.015033778217103748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,1536,0.0086986662613021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,1536,0.012808000048001608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,1024,0.008050666915045844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,6144,0.02139555580086178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,768,0.003867555409669876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,1024,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,2560,0.01684888866212633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,3584,0.014868444866604276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,512,0.003556444413132138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,768,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,768,0.0069822221994400024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,768,0.010067555639478896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,256,0.0032008888406885993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,512,0.007318221860461765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,3072,0.018568888306617737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,256,0.005579555614127054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,128,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,1024,0.011139555937714048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,128,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,128,0.008360889222886827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,1536,0.0090524446633127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,512,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,32,0.0031884445084465873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,256,0.007299555672539606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,64,0.005218666460778978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,256,0.008012444608741337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,65536,0.05345511105325487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,51200,0.04351288742489285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,32,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,16384,0.016639999217457242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,12288,0.013456000222100152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,65536,0.1886533366309272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,51200,0.14795555008782282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,128,0.007248000138335758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,1024,0.0042444442709287005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,12288,0.03776622149679396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,7168,0.010268444816271463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,10240,0.032454222440719604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,6144,0.009349333743254343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,8192,0.026761778526835974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,7168,0.023935111032591924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,6144,0.021651556094487507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,4096,0.007982222570313348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,5120,0.01864177816443973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,10240,0.012105777859687805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,3584,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,512,0.006232000059551663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,4096,0.016192000773217943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,2560,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,3072,0.01346577786737018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,2048,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,5120,0.009462222456932068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,1536,0.004910222358173794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,2048,0.010729778144094678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,1024,0.003916444463862313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,1536,0.00908088849650489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,64,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,1024,0.007720000214046902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,512,0.0032266666077905228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,768,0.00702311098575592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,256,0.0032506665835777917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,2560,0.012085333466529846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,512,0.006228444476922353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,128,0.0028479999552170434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,64,0.0025519999778932994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,256,0.005581333405441708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,16384,0.050943109724256724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,8192,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,64,0.005202666752868229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,65536,0.04929955469237434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,32,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,51200,0.04153955645031399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,65536,0.18756977717081705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,16384,0.010403555300500657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,51200,0.14766844113667807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,12288,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,768,0.0038853333228164246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,16384,0.0500257776843177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,3584,0.014555555250909595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,8192,0.007676444119877285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,12288,0.03801511062516107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,7168,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,10240,0.03222577770551046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,6144,0.007696888513035244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,5120,0.00701422244310379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,8192,0.026713778575261433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,4096,0.014821334017647637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,6144,0.02144444485505422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,3584,0.013807111316257052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,5120,0.018944000204404194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,128,0.004980444494220945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,3072,0.011762667033407422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,4096,0.016399110356966656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,2560,0.010418666733635796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,2048,0.005232000102599462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,3584,0.014750222365061441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,1536,0.004578666554556953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,2560,0.012094222009181976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,1024,0.003959111041492886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,2048,0.010716444088353051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,768,0.003882666842805015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,1536,0.009027555584907532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,512,0.003552000141806073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,1024,0.007672888537247975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,256,0.003192000091075897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,3072,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,512,0.006061333335108227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,128,0.002826666666401757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,64,0.002564444517095884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,256,0.005495111147562663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,32,0.0028888889484935333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,128,0.005084444251325396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,64,0.004984000076850255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,32,0.004943999979231092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,7168,0.02420888841152191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,768,0.006999111009968652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,32,0.002893333426780171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,10240,0.00830666638082928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,3072,0.013473778135246701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,16384,0.9191812939114041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,16384,2.6857645246717663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,12288,0.6717066764831543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,12288,2.0898284912109375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,16384,2.96012454562717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,10240,0.5613226890563965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,51200,3.2770426008436413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,10240,1.834731631808811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,8192,1.3768995073106554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,10240,1.8477662404378254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,8192,0.45741600460476345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,8192,1.360380490620931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,7168,0.41344534026251895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,7168,1.2078506681654189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,7168,1.1503057479858398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,6144,0.35505066977606875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,12288,2.2782185872395835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,5120,0.8527706464131674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,6144,0.9869457880655924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,5120,0.3036737706926134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,4096,0.6707555452982584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,4096,0.2514737712012397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,5120,0.8137502140469022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,6144,1.1141751607259114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,3584,0.7495857874552408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,4096,0.6628230942620171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,3072,0.4812773068745931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,3584,0.5762782096862793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,3072,0.25000977516174316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,2560,0.4982542461819119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,2560,0.170141339302063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,2560,0.44117244084676105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,2048,0.4697599940829807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,2048,0.14355644914839003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,1536,0.2374337779151069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,2048,0.33890843391418457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,51200,8.714899698893229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,1024,0.16701689031389025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,1536,0.11470133728451198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,1536,0.36166222890218097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,1024,0.20737866560618082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,3072,0.493468443552653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,768,0.08174310790167914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,1024,0.09034399853812324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,256,0.07461066378487481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,512,0.07408177852630615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,512,0.14041688707139757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,3584,0.22287911838955352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,256,0.0666382246547275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,256,0.12524710761176214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,768,0.139500445789761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,128,0.06403022342258029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,64,0.05137777659628126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,32,0.053767111566331655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,768,0.16249688466389975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,64,0.119851549466451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,51200,12.194386800130209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,128,0.06009155511856079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,128,0.11798311604393853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,512,0.10615466700659858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,32,0.1202933390935262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,16384,2.0327991909450955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,16384,2.325768788655599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,51200,2.596168941921658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,65536,3.4394124348958335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,12288,1.509897761874729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,12288,0.5159404542711046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,16384,0.6645573510064019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,12288,1.7256168789333766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,10240,0.4360231028662787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,10240,1.3426986270480687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,10240,1.3792969385782878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,8192,1.1720497343275282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,8192,0.35796533690558535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,51200,6.425812615288629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,8192,1.057785775926378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,7168,0.8549093140496148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,7168,0.3206088807847765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,6144,0.7282186614142524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,6144,0.27947645717196995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,7168,0.9255315992567273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,6144,0.7679351170857748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,5120,0.23652532365587023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,4096,0.5172728962368435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,5120,0.6748799747890897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,51200,9.504146152072483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,3584,0.49968269136216903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,65536,8.240587022569445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,3584,0.17578932974073622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,3584,0.4416862328847249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,3072,0.15710755189259848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,2560,0.2962346606784397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,4096,0.1954044500986735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,2560,0.13757510979970297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,2048,0.25499023331536186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,5120,0.6587297651502821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,3072,0.38434844546847874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,1536,0.22294044494628906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,2048,0.2672399944729275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,1024,0.13745778136783177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,1536,0.21229155858357748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,2560,0.3271733389960395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,1024,0.07213511069615682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,1024,0.15894933541615805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,65536,12.281327141655815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,512,0.08010577493243747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,768,0.1280168824725681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,1536,0.09206577804353501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,256,0.05815822548336453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,4096,0.5072684288024902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,512,0.11110933621724446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,128,0.047503110435273915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,256,0.05412800113360087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,256,0.0994044476085239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,64,0.04109777675734626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,2048,0.11796000268724229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,768,0.06479733520083956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,64,0.0948408908314175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,512,0.05942133400175306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,32,0.09540000226762559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,3072,0.38346221711900497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,768,0.11099288860956828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,128,0.05201244354248047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,128,0.09335822529262966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,32,0.04166577921973334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,16384,0.5989626778496636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,65536,1.0204310946994357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,51200,0.775785764058431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,16384,0.22819643550448945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,16384,0.7329413096110026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,12288,0.575934198167589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,65536,2.6448035769992404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,51200,2.279984792073568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,10240,0.43389156129625106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,12288,0.5305182139078776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,10240,0.15147733688354492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,10240,0.4675137731764052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,8192,0.1204799943500095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,7168,0.30111998981899685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,7168,0.1085946692360772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,6144,0.2414604557885064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,65536,3.6662398444281687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,5120,0.2239075501759847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,6144,0.2551848888397217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,6144,0.0963840021027459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,8192,0.35029954380459255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,5120,0.08130311303668551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,4096,0.18835200203789604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,5120,0.21741155783335367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,4096,0.16954399479760063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,7168,0.29865066210428876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,3584,0.061668442355261914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,3584,0.14932178126441106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,3072,0.1452088885837131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,3072,0.05486311184035408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,3072,0.13108711772494847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,2560,0.1256728834576077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,12288,0.1753013398912218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,2560,0.04851288927925957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,2048,0.0909786687956916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,2560,0.112500442398919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,1536,0.06576889091067843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,2048,0.04186044467820061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,2048,0.09277511305279201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,1536,0.03399822115898132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,1024,0.046462221278084644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,1536,0.07516888777414958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,1024,0.027815110153622095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,1024,0.0565226674079895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,512,0.026327111654811438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,3584,0.1474399964014689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,768,0.045313778850767344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,256,0.02097333305411869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,512,0.02351911034848955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,512,0.037645333343082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,128,0.017685333887736004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,256,0.033921778202056885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,8192,0.4041751225789388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,128,0.03226577904489305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,128,0.02086044516828325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,32,0.01576266686121623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,768,0.02544266647762722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,4096,0.0687848859363132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,64,0.0321413344807095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,32,0.03230133321550157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,256,0.021175111333529156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,768,0.036306666003333196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,51200,2.977420383029514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,65536,0.7406613561842176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,16384,0.4947937859429254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,51200,0.5938266648186578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,65536,1.9883164299858942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,16384,0.18085155222151014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,16384,0.5364746517605251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,64,0.015139554937680563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,12288,0.1396284500757853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,10240,0.3142782317267524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,51200,2.0465965270996094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,10240,0.11940356095631917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,65536,2.80060174730089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,12288,0.3987297746870253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,8192,0.23950311872694227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,8192,0.09656000137329102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,7168,0.23263555102878145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,51200,1.8051092359754775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,6144,0.2236835426754422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,8192,0.2634035481346978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,7168,0.22876445452372232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,5120,0.17676356103685167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,6144,0.19995555612776014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,5120,0.06274400154749553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,4096,0.145395557085673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,5120,0.16397778193155924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,3584,0.1300897730721368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,10240,0.3284337785508898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,4096,0.13411733839246961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,3072,0.10134044620725845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,3584,0.11577067110273574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,7168,0.08316177792019315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,2560,0.0853288902176751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,3072,0.10182044241163467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,12288,0.3979626761542426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,2048,0.06157333321041531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,6144,0.07273244195514254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,1536,0.054021331999037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,2048,0.03163911236657037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,2048,0.0717466672261556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,4096,0.052229334910710655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,1536,0.026263111167483862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,1536,0.05848177936342028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,3584,0.05029155479537117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,768,0.027075555589463975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,1024,0.02221422228548262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,1024,0.04355022311210632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,3072,0.04184622234768338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,768,0.020273778173658583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,256,0.017621333400408428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,512,0.01846311158604092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,2560,0.037307557132509016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,512,0.029118223322762385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,128,0.01406133340464698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,256,0.016812443733215332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,64,0.01219466659757826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,256,0.02623111175166236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,128,0.01688799924320645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,32,0.01254755589697096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,128,0.02510044475396474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,64,0.025055999557177227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,32,0.024715556038750544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,512,0.02199288871553209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,768,0.035359111097123884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,2560,0.0870480007595486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,1024,0.034846223062939115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,65536,0.6510115729437934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,51200,1.2454338073730469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,51200,0.49331288867526585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,65536,1.8812550438774958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,12288,0.3190488815307617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,16384,0.6227315796746148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,12288,0.3392595450083415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,12288,0.11488266785939534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,10240,0.28787732124328613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,65536,2.5636683570014105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,10240,0.09804799821641709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,16384,0.4563351207309299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,10240,0.27123112148708767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,8192,0.07913066943486531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,8192,0.23950399292839897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,7168,0.07101244396633573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,6144,0.16962488492329916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,7168,0.19320978058709037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,5120,0.12923288345336914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,6144,0.06212355693181356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,51200,1.7484187020195856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,5120,0.13935289118025038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,8192,0.23015822304619682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,4096,0.04439466529422336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,4096,0.11134400632646348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,16384,0.15222756067911783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,3584,0.04050488935576545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,3072,0.07333066728379992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,3584,0.09881244765387641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,2560,0.058556444115108915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,3072,0.03563288847605387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,6144,0.1675431066089206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,2048,0.051310221354166664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,2560,0.07418844434950086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,3584,0.1017200019624498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,7168,0.20596888330247667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,2048,0.06159910890791151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,2048,0.02797422144148085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,1024,0.030165334542592365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,1536,0.023210666245884363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,1536,0.04933422141604953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,768,0.021663111117151048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,4096,0.11444800429873997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,2560,0.03183733423550924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,512,0.0174346665541331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,768,0.030608889129426744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,1536,0.042308443122439914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,256,0.01354311075475481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,512,0.025349333882331848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,256,0.015145777000321282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,128,0.012038222617573209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,256,0.022456889351209004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,5120,0.053671999110115894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,64,0.010254222485754225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,128,0.021664889322386846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,768,0.017861333158281114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,32,0.010374222364690568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,64,0.02168888847033183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,32,0.021321778496106465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,1024,0.019816888703240287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,3072,0.0861404471927219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,128,0.015156444576051501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,65536,0.5071679751078287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,65536,1.2831830978393555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,512,0.01623288955953386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,16384,0.4051351017422146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,51200,0.4094080130259196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,16384,0.12598310576544866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,51200,1.1131466759575737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,16384,0.3774275514814589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,65536,1.7762755288018122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,12288,0.09684089157316421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,1024,0.0376800000667572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,8192,0.18323377768198648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,51200,1.4662115308973525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,10240,0.22461244795057508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,8192,0.06656710969077216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,8192,0.17972889211442736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,6144,0.11845244301689996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,7168,0.15785333845350477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,12288,0.2701164351569282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,6144,0.05204622281922234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,6144,0.13627644379933676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,12288,0.34685778617858887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,5120,0.04335911075274149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,4096,0.09300000137752956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,5120,0.11528978082868789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,7168,0.15961066881815592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,3584,0.08204178015391032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,7168,0.05973066886266073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,3584,0.032961779170566134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,3072,0.07204000155131023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,3584,0.08101244767506917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,3072,0.02979466650221083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,3072,0.07153599792056613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,10240,0.08094577656851874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,2560,0.027447111076778833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,2560,0.06160889069239298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,2048,0.04408444298638237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,4096,0.09222310781478882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,1536,0.035559111171298556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,2048,0.05132800009515551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,2560,0.057067553202311196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,1536,0.020067556036843192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,1024,0.024875554773542616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,1536,0.04115466607941522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,768,0.018953777021831937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,1024,0.016951999730534024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,1024,0.03170488940344916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,512,0.015221332510312399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,4096,0.036531554328070745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,768,0.025840888420740765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,256,0.011744889120260874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,512,0.014358222484588623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,2048,0.023781332704755995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,512,0.02141333288616604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,5120,0.1179013384713067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,256,0.013443555268976422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,64,0.008823110825485652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,128,0.013462222284740873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,128,0.018464888135592144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,32,0.009365333451165093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,64,0.018280888597170513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,32,0.018035555879275005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,128,0.010054222411579555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,10240,0.24837422370910645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,256,0.018964444597562153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,65536,0.48987287945217556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,65536,1.2553733189900715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,16384,0.25979288419087726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,768,0.015458666616015963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,12288,0.18776267104678682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,16384,0.3207520114050971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,65536,1.6423288981119792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,51200,1.1919938193427193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,10240,0.19853688610924614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,12288,0.23448265923394096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,10240,0.07340088817808363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,10240,0.19809422228071424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,8192,0.0585119989183214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,51200,0.37842843267652726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,7168,0.13053422504001194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,51200,0.9693466822306315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,6144,0.10971911085976495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,7168,0.052095111873414784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,7168,0.13933600319756403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,12288,0.08402044243282741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,6144,0.12039289209577773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,5120,0.04005600015322367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,8192,0.13098133934868708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,5120,0.1020044419500563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,4096,0.032794667614830866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,8192,0.15814310974544948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,4096,0.08248533142937554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,16384,0.13194666968451604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,3072,0.0585377746158176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,3584,0.030011554559071858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,2560,0.05519022213088142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,3584,0.07300000058280097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,3072,0.06458222203784518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,2560,0.02494222256872389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,4096,0.08295200268427531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,2560,0.055161777469846934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,1536,0.029417776399188574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,2048,0.02164622147878011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,3584,0.05919466416041056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,2048,0.0463048881954617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,1024,0.02252888845072852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,1536,0.01862044466866387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,1536,0.03733333283000522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,1024,0.01589866644806332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,1024,0.028374221589830186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,512,0.013836444252067141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,768,0.01440533333354526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,768,0.023777777949968975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,256,0.010429333481523724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,2048,0.03704799877272712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,512,0.019493333167499967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,128,0.009711999860074785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,256,0.01720888912677765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,128,0.01258044441541036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,128,0.016788444585270353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,64,0.008664000365469191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,32,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,5120,0.10369688934750027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,64,0.0165964447789722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,512,0.013334222137928009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,256,0.012691555751694573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,768,0.01732711162832048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,65536,1.030944930182563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,6144,0.04628622200753954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,51200,0.8711075782775879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,32,0.016536888149049547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,51200,0.3380124568939209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,16384,0.09933688905504014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,16384,0.3385599984063043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,12288,0.20791289541456434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,65536,0.39752711190117734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,10240,0.17796622382269967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,12288,0.2081608904732598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,16384,0.26775911119249135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,65536,1.3394222259521484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,8192,0.14480711354149714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,8192,0.0512951115767161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,8192,0.13993067211574978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,7168,0.1261448860168457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,12288,0.07916000154283312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,7168,0.04541155695915222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,6144,0.10919644435246785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,6144,0.040440890524122454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,6144,0.10592710971832275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,5120,0.08756533596250747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,5120,0.03456711106830173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,51200,1.0884560479058158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,5120,0.08925422032674153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,4096,0.06913422213660346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,3584,0.05713422430886162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,3072,0.027136888768937852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,4096,0.02971822354528639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,3072,0.05092089043723213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,7168,0.12338044908311631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,3584,0.06428444385528564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,3072,0.02444533341460758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,10240,0.06350933180914985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,3072,0.05667644739151001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,2048,0.035397334231270686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,2560,0.022264889544910852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,2560,0.04899555444717407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,1536,0.026745777991082933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,2048,0.019739554988013375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,1024,0.02033511135313246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,1536,0.01718577742576599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,1536,0.03351199958059523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,1024,0.014537778165605334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,1024,0.02609155575434367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,4096,0.07243555784225464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,512,0.012330667012267642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,768,0.013454222016864352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,768,0.021200888686709937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,256,0.010042666561073726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,512,0.017481777403089736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,2048,0.041294223732418485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,128,0.008559111091825698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,768,0.0159262220064799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,3584,0.02681066592534383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,128,0.014896889527638754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,128,0.011744889120260874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,32,0.007576000359323289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,10240,0.1714408927493625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,512,0.012436444560686747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,64,0.014928888943460254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,32,0.014860444598727755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,256,0.011690666278203329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,256,0.01554666625128852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,65536,0.36857954661051434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,65536,0.8196738031175402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,64,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,51200,0.2873866558074951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,16384,0.2337404357062446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,16384,0.08250844478607178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,65536,1.1344470977783203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,16384,0.23304088910420737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,10240,0.14068622059292263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,10240,0.0632657806078593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,10240,0.1480791171391805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,8192,0.11189244853125678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,2560,0.0444160004456838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,51200,0.7701750861273872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,8192,0.0463137792216407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,8192,0.11908177534739177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,12288,0.17711910936567518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,7168,0.03987110985649957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,51200,0.8521155781216092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,7168,0.10536799828211467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,12288,0.06395822101169162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,6144,0.0353999998834398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,6144,0.09158133135901557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,5120,0.07738666401969062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,4096,0.06241689125696818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,7168,0.08258133464389376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,3584,0.048581331968307495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,6144,0.08662133084403144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,3584,0.02385866641998291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,3072,0.041125333971447416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,12288,0.1844471163219876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,2560,0.036460445986853704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,3072,0.022016000416543748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,4096,0.05735911263359917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,3072,0.04900888933075798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,2048,0.027738667196697656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,2560,0.020447111792034574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,1536,0.02200977835390303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,2048,0.017811555001470778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,4096,0.026015111141734656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,2048,0.035699556271235146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,1024,0.016166221764352586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,1536,0.015507555670208402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,1536,0.029137776957617864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,768,0.01313866674900055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,3584,0.05544888973236084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,1024,0.022104889154434204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,512,0.010093332992659675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,5120,0.030625777112113103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,768,0.018983110785484314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,256,0.008354666332403818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,512,0.011639110743999481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,2560,0.04257511099179586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,512,0.015511110424995422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,128,0.007244444555706448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,256,0.011404444773991903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,64,0.006618666566080517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,256,0.013864888913101621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,32,0.00666133314371109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,128,0.013445333474212222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,1024,0.013470222552617392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,5120,0.0711688862906562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,32,0.013258666627936892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,768,0.012441777520709567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,65536,0.5574426651000977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,65536,0.3026328881581624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,51200,0.4687217606438531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,16384,0.18006222777896455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,128,0.011063999599880643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,65536,0.876223087310791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,12288,0.1350933313369751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,51200,0.7017146746317545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,12288,0.05505511164665222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,10240,0.12094667222764756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,12288,0.15671822759840223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,8192,0.10091555780834621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,10240,0.047001779079437256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,10240,0.12389777766333686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,7168,0.08516089121500652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,8192,0.03791111045413547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,6144,0.07183555761973064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,7168,0.03404444456100464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,51200,0.23292356067233613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,7168,0.088664001888699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,6144,0.03036710951063368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,16384,0.19324621889326307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,5120,0.02681955529583825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,5120,0.06526133086946276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,4096,0.02315644423166911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,4096,0.053393777873780995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,8192,0.10001333554585774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,3584,0.02107911143038008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,16384,0.08432977729373509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,3072,0.03641866644223531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,64,0.01348444405529234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,4096,0.0447671115398407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,2560,0.028590222199757893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,3072,0.04221866528193156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,2048,0.02347022294998169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,2560,0.03714222378200955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,3584,0.04186488853560554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,1536,0.019660444723235238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,2048,0.015570667054918079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,2048,0.031120889716678198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,5120,0.05867377916971842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,1024,0.014535110857751636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,1536,0.024868445263968572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,6144,0.07613244321611193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,768,0.011920889218648275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,1024,0.012147555748621622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,512,0.009441778063774109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,1024,0.019279999865425956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,3072,0.019696000549528334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,768,0.016120000018013846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,512,0.010477333433098264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,2560,0.017401778035693698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,512,0.01385600037044949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,256,0.010054222411579555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,256,0.012087110843923358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,128,0.010052444206343757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,1536,0.013819555441538492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,32,0.005975111077229182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,64,0.011774222056070963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,768,0.01110755569405026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,32,0.011746666497654386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,256,0.007287999822033777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,128,0.006576889091067844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,65536,0.48192797766791445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,65536,0.28047998746236164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,51200,0.47528802023993594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,16384,0.1564977831310696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,51200,0.2263582282596164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,16384,0.07616977559195624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,51200,0.6268533070882162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,16384,0.2231848769717746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,10240,0.09942577944861518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,12288,0.1321315500471327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,10240,0.04657511247528923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,64,0.00592533333433999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,10240,0.11256444454193115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,65536,0.793040010664198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,12288,0.11592888832092285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,12288,0.05477777785725064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,3584,0.047930667797724404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,128,0.01178311142656538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,7168,0.033955554167429604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,8192,0.0366888874106937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,6144,0.06926311386956109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,5120,0.05224800109863281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,7168,0.0803697771496243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,5120,0.05881066454781426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,8192,0.08054222000969781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,4096,0.0396106673611535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,6144,0.06281599733564588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,3584,0.03496444556448195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,4096,0.022280888424979314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,4096,0.0483457777235243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,8192,0.08948177761501735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,3072,0.03009422289000617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,7168,0.06711377700169881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,3584,0.04342044393221537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,2560,0.026175111532211304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,2048,0.02204888893498315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,2560,0.016826666063732572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,2560,0.03329688972897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,2048,0.015206222732861837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,2048,0.028357333607143823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,5120,0.025928000609079998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,1024,0.013510222236315409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,1536,0.013471999930010902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,1536,0.02257866660753886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,768,0.011160000330872007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,1024,0.01180622229973475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,1024,0.017621333400408428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,512,0.009355555805895064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,6144,0.029438220792346533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,768,0.011023999916182624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,3072,0.01869066721863217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,256,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,3072,0.038352890147103205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,128,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,512,0.012537777423858643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,256,0.00981866651111179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,1536,0.017981333865059745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,64,0.005588444570700328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,32,0.00565244464410676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,128,0.01072533345884747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,64,0.01074577785200543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,3584,0.02015555567211575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,65536,0.5438248846266005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,768,0.01516355574131012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,512,0.010377777947319878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,65536,0.25830578804016113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,16384,0.13626044326358372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,51200,0.4786835776435004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,51200,0.20042310820685494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,12288,0.10673955414030288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,16384,0.08096088965733846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,16384,0.15596177842881945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,10240,0.09709066814846462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,12288,0.11755732695261638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,10240,0.04332977864477369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,10240,0.09869333108266194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,256,0.011420444481902652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,65536,0.6861119800143771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,7168,0.06521955463621351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,8192,0.08002399735980563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,51200,0.5372728771633571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,32,0.010760000182522668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,6144,0.05793866846296522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,7168,0.03127288818359375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,12288,0.05714044306013319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,7168,0.07122755712933011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,6144,0.028314666615592107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,4096,0.03730222251680162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,6144,0.06173511346181234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,5120,0.05211822191874186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,8192,0.07832622528076172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,128,0.009761778016885122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,4096,0.04288977715704176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,8192,0.03528088993496365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,3584,0.01906844476858775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,3584,0.0388302206993103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,2560,0.024152888192070857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,3072,0.01744622157679664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,3072,0.034480889638264976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,5120,0.04725422130690681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,2560,0.01606399979856279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,2560,0.02981333269013299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,5120,0.024784889486100938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,1536,0.0166275550921758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,2048,0.025472000241279602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,4096,0.021566222111384075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,1024,0.012607999973826937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,1536,0.020576889316240948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,3072,0.028029332558314007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,768,0.010378667049937779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,1024,0.016026667422718473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,768,0.010432889064153036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,768,0.013796444568369122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,2048,0.020394666327370536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,512,0.011729777687125735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,2048,0.01424888935354021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,3584,0.03281689021322463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,1536,0.012201777762836881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,128,0.005577777822812398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,256,0.009416888985368941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,256,0.010372444159454769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,64,0.005236444373925527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,128,0.009272889130645329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,512,0.008301333420806462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,64,0.010053333308961656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,512,0.009868444667922126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,256,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,65536,0.44744178983900285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,1024,0.011056889262464313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,32,0.005586666779385672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,65536,0.5811431143018935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,16384,0.11902488602532281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,51200,0.17959111266665986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,51200,0.4549262258741591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,12288,0.09442489014731513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,16384,0.05646666553285387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,16384,0.13573511441548666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,12288,0.043431109852261014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,12288,0.10188888841205174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,128,0.010036444498433007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,8192,0.07113866673575507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,8192,0.030881779061423406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,32,0.010021333065297868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,8192,0.0703164471520318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,7168,0.060573332839541964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,6144,0.05127288897832235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,7168,0.027229333917299908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,7168,0.06237066454357571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,10240,0.0825164450539483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,5120,0.04191911220550537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,6144,0.054082665178510875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,10240,0.03801066676775614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,5120,0.021749332547187805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,5120,0.0462097790506151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,3584,0.02878311276435852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,4096,0.018626666731304593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,4096,0.03784888982772827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,3072,0.025252444876564875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,3584,0.03442400031619602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,6144,0.024705777565638225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,3072,0.03017688790957133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,2560,0.02141600019401974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,4096,0.033654222885767617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,10240,0.08602844344245063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,2560,0.014503111441930136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,2560,0.026122666067547266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,1536,0.014965333872371249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,2048,0.013292444248994192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,2048,0.02211911148495144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,3584,0.017272889614105225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,1536,0.0120000003112687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,1536,0.01829688913292355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,768,0.009390222529570261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,1024,0.01036444471942054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,65536,0.23263290193345812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,1024,0.01424888935354021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,512,0.007416889071464539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,51200,0.38271021842956543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,2048,0.01792355544037289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,512,0.009390222529570261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,512,0.010445333189434476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,128,0.005591111050711737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,256,0.009020444419648912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,256,0.009682666924264696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,1024,0.011211555865075854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,64,0.0052355556852287715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,32,0.005418666534953647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,3072,0.015962666935390897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,64,0.009115555220180089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,768,0.009833777944246927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,32,0.009019555317031013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,256,0.005968889014588461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,65536,0.3204702271355523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,65536,0.1702951060401069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,51200,0.3053964508904351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,16384,0.09500710831748115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,51200,0.13422400421566433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,65536,0.49339909023708767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,128,0.008993777963850234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,16384,0.04591022266281975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,51200,0.3902435567643907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,10240,0.06119733386569553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,12288,0.03439111179775662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,12288,0.08971199724409316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,10240,0.03053155541419983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,768,0.012295111185974546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,10240,0.07618933253818087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,7168,0.04239999916818407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,8192,0.02404444416364034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,6144,0.03710755705833435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,8192,0.061943113803863525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,7168,0.021910222040282354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,5120,0.03112799922625224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,12288,0.07001066870159574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,6144,0.01974311139848497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,5120,0.017601778109868366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,8192,0.04911111129654778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,4096,0.023843554986847773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,4096,0.015789333317014907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,3584,0.020920000142521326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,7168,0.05476977758937412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,3584,0.031011554929945204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,3584,0.014148443937301636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,3072,0.01865244408448537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,3072,0.01331733332739936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,2560,0.016315556234783597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,3072,0.027202667461501226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,2560,0.012147555748621622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,2560,0.02372177773051792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,2048,0.0114702218108707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,128,0.009317333499590555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,5120,0.04087111022737291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,1536,0.016693333784739178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,1024,0.009000889129108852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,2048,0.013595555391576556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,1024,0.009408888717492422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,768,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,2048,0.020115555988417733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,6144,0.047822223769293896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,512,0.006640888750553131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,768,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,768,0.011234666738245221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,256,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,512,0.008665777742862701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,512,0.009712888962692684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,128,0.004940444396601783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,256,0.008261333737108443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,256,0.008731555607583787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,64,0.00434400017062823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,128,0.008264888491895463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,32,0.004226666771703296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,16384,0.12143911255730523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,64,0.00832622249921163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,32,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,65536,0.3048746585845947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,1536,0.010263111028406356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,65536,0.15450399451785618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,51200,0.23588710361056855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,65536,0.39696444405449766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,16384,0.07553511195712619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,1024,0.013117333253224691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,128,0.008360000120268928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,16384,0.043712887499067515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,51200,0.3085235489739312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,16384,0.100399997499254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,10240,0.047154668304655284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,12288,0.03405866689152188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,8192,0.03750400079621209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,10240,0.029885331789652508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,10240,0.06304622358746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,8192,0.023553777072164748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,4096,0.03405244482888116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,8192,0.05162400007247925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,6144,0.02709688908523983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,7168,0.02130399975511763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,7168,0.04581244455443489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,12288,0.057109336058298744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,6144,0.019139554765489366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,1536,0.011519999967681037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,4096,0.01975733372900221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,5120,0.01704355577627818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,12288,0.07483555873235066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,5120,0.0345164438088735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,3584,0.018237334158685472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,4096,0.015003555350833468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,3072,0.01601777805222405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,3584,0.013938667045699226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,3584,0.025809778107537165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,3072,0.013008889224794177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,51200,0.12312977843814427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,3072,0.023119999302758112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,2048,0.012191111014948951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,5120,0.024028443627887305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,2560,0.012095111111799875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,6144,0.04032533367474874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,2048,0.010932444698280759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,2048,0.0173333336909612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,1024,0.007979555262459649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,1536,0.014899555179807874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,7168,0.03242399957444932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,768,0.007106666763623555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,1024,0.011617778076065911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,2560,0.014477334088749357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,512,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,768,0.010038221875826517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,1536,0.010038221875826517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,512,0.008515555825498369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,512,0.008692444198661381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,1536,0.010075555907355415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,256,0.008059555457697975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,128,0.004942222187916438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,256,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,64,0.004563555535342959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,128,0.008022222254011366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,128,0.007300444775157505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,32,0.004874666531880696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,64,0.007358222372002072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,2560,0.020224000016848247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,256,0.00490133340160052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,65536,0.17358933554755318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,32,0.007606222397751278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,51200,0.13574666447109646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,65536,0.1102551089392768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,4096,0.028943998946083918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,65536,0.3030150996314155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,51200,0.08920177486207749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,51200,0.23630134264628092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,12288,0.03900266687075297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,16384,0.032060444355010986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,768,0.008848889006508721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,16384,0.07833155658509997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,12288,0.02455111179086897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,8192,0.0247653325398763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,12288,0.05905510981877645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,10240,0.022452443838119507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,7168,0.022806222240130108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,8192,0.0174666659699546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,6144,0.019832889238993328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,8192,0.0417848891682095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,7168,0.01609422266483307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,16384,0.0517813331551022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,7168,0.03694044550259908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,1024,0.009352000223265754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,4096,0.014451555079884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,6144,0.03268622358640035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,5120,0.02790844440460205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,10240,0.03408177693684896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,4096,0.011807999677128263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,4096,0.02351288828584883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,3072,0.011682666838169098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,3584,0.011910222470760345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,10240,0.05043466554747688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,2560,0.010477333433098264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,3072,0.010946667028797997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,3072,0.019132445255915325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,2048,0.009299555586444007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,2560,0.01015377789735794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,6144,0.014730667074521383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,2560,0.01666933298110962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,1536,0.007809778054555257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,5120,0.01348355578051673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,1024,0.006292444550328785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,3584,0.013139555851618448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,1536,0.012088889049159156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,768,0.005631999837027655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,1024,0.008471999731328752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,1024,0.010027555955780877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,512,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,768,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,768,0.00872444444232517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,256,0.004234666625658671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,512,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,512,0.007703111403518253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,128,0.0042088888585567474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,256,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,2048,0.01447111037042406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,64,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,128,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,128,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,32,0.004233777936961916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,3584,0.02103288968404134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,65536,0.15929599603017172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,32,0.006682666639486949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,5120,0.01720888912677765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,51200,0.12534666061401367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,65536,0.26105599933200413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,256,0.007359111474619971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,16384,0.04319644305441114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,1536,0.00904622260067198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,12288,0.030457778109444514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,16384,0.06878666745291816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,64,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,12288,0.0236435549126731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,10240,0.02388266722361247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,65536,0.10995288689931233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,10240,0.022340445054901972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,8192,0.020260444945759244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,10240,0.04383288820584615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,7168,0.018567999203999836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,51200,0.09066222111384074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,8192,0.03624799847602844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,6144,0.01575555569595761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,16384,0.031870222753948636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,7168,0.0324577788511912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,5120,0.01423822177780999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,6144,0.015167999598715039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,6144,0.02863822380701701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,4096,0.011727111207114326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,5120,0.024193776978386775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,2048,0.009702222214804756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,4096,0.012597333225939008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,7168,0.017269333203633625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,51200,0.204912887679206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,3584,0.0188719994492001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,3584,0.01112711098459032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,12288,0.05169777737723457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,5120,0.013260444833172692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,2560,0.008838222258620793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,3584,0.011239999698268043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,2560,0.014855111638704935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,2048,0.00794577764140235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,2560,0.010398222340477837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,2048,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,1536,0.0069520001610120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,3072,0.017064000169436138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,1536,0.01093777765830358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,1024,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,1536,0.00904888908068339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,8192,0.01737422247727712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,768,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,1024,0.008383110993438298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,1024,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,512,0.004558222161398994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,768,0.008031999899281396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,3072,0.010035555395815108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,256,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,512,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,512,0.006959999601046245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,128,0.004110222061475118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,256,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,4096,0.020787555310461257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,64,0.00388355553150177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,128,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,32,0.0038888889054457345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,65536,0.12074222829606797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,64,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,768,0.008060444560315874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,2048,0.01312622262371911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,51200,0.09799822171529134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,256,0.007336888876226213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,51200,0.0853288902176751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,51200,0.17181244161393908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,128,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,16384,0.02796622117360433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,16384,0.05683466460969713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,32,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,12288,0.04378844300905863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,10240,0.023852444357342188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,65536,0.10566400157080756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,8192,0.019688889384269714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,65536,0.2178408834669325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,10240,0.03737955623202854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,16384,0.03916533456908332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,7168,0.01754755609565311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,8192,0.01663822266790602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,12288,0.028007109959920246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,6144,0.015607110328144498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,7168,0.02807377775510152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,5120,0.013648889131016202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,6144,0.013814222481515674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,12288,0.022700443863868713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,6144,0.02444444431198968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,4096,0.011687999798191918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,5120,0.012497777740160624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,5120,0.02132266594303979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,3584,0.010374222364690568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,3072,0.01107377807299296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,4096,0.01793688866827223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,3072,0.009393778112199571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,3584,0.0107342220014996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,2560,0.008664000365469191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,8192,0.03122755553987291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,3072,0.014838222000333997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,7168,0.01498311095767551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,2560,0.009741333623727163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,2560,0.013181333740552267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,1536,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,2048,0.009353777600659264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,2048,0.011448000040319232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,1024,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,10240,0.020943111843532983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,1536,0.010081777969996134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,768,0.004579555657174853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,3584,0.016556445095274184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,1024,0.008168000313970778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,512,0.004228444563017951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,768,0.007857778006129796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,2048,0.0074275558193524676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,256,0.00388711111413108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,512,0.007336888876226213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,512,0.006506666541099548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,128,0.0035537779331207275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,1536,0.008664000365469191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,4096,0.011115555961926779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,64,0.0033697777738173804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,128,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,32,0.003318222239613533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,1024,0.008344000412358178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,64,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,65536,0.07841244671079847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,32,0.00563822231358952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,51200,0.06276710828145345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,256,0.007356444166766272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,256,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,128,0.007238222493065729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,51200,0.13927733898162842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,3072,0.010382222632567087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,768,0.007342221836249034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,16384,0.0468426677915785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,12288,0.01886488828394148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,65536,0.10514133506351048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,65536,0.17284799946679008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,51200,0.08465866910086738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,10240,0.030541333887312148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,8192,0.013450666434235044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,10240,0.020512888828913372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,16384,0.027109333210521277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,8192,0.016136000553766888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,7168,0.012112889024946423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,12288,0.02234488891230689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,7168,0.014945778581831189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,6144,0.010886222124099731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,7168,0.02269777821169959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,5120,0.009396444592210982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,6144,0.013816888961527081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,6144,0.02053244411945343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,4096,0.008061333662933773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,5120,0.017857778403494094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,3584,0.007635555333561367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,4096,0.015375999940766228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,3584,0.014144889182514615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,3072,0.007080000307824876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,16384,0.02275200022591485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,3072,0.01279022213485506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,5120,0.012454222473833295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,2560,0.006324444380071428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,3072,0.010120000276300643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,4096,0.011160000330872007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,2560,0.009702222214804756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,2560,0.011620444556077322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,1536,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,3584,0.010729778144094678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,10240,0.01612977849112617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,1024,0.00424888895617591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,1536,0.00871022211180793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,1536,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,768,0.0038746665749284956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,1024,0.008048888709810045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,1024,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,512,0.003530666646030214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,768,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,8192,0.025587555434968736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,2048,0.009360888765917884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,512,0.005986666513813867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,2048,0.010334221853150262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,12288,0.035522666242387556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,256,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,128,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,64,0.003128000017669466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,768,0.007705777883529663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,2048,0.005584888988071018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,32,0.0031893334041039148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,65536,0.05186400148603651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,256,0.0035413333939181436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,32,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,128,0.0031759999692440033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,65536,0.1047262218263414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,65536,0.1724239985148112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,16384,0.01661777827474806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,51200,0.08437866634792751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,128,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,51200,0.13559556007385254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,16384,0.026533333791626826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,16384,0.0460382236374749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,512,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,64,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,256,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,10240,0.030132446024152968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,12288,0.013961777091026306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,10240,0.012811555630630918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,12288,0.03533777925703261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,7168,0.010984000232484607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,51200,0.042452444632848106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,8192,0.011419555379284753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,7168,0.014863999353514777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,7168,0.022610666023360357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,5120,0.009379555781682333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,8192,0.025032000409232244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,5120,0.012472000386979846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,5120,0.017663111289342243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,4096,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,8192,0.015999999311235216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,3584,0.0075297777851422625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,4096,0.011065777805116443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,10240,0.020842666427294414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,4096,0.015206222732861837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,6144,0.010426667001512317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,3584,0.01074577785200543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,2560,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,3584,0.014072888427310519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,6144,0.013776000175211163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,3072,0.012838222086429596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,6144,0.020073778099483915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,2560,0.011573333707120685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,12288,0.02204799983236525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,2048,0.010097777677906884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,1536,0.0048942222363419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,1536,0.008692444198661381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,1024,0.004223111189073986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,3072,0.010095111197895473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,1024,0.007303111255168915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,2048,0.005586666779385672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,768,0.003891555385457145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,768,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,2048,0.00924800005224016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,768,0.006923555499977536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,512,0.007333333293596904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,256,0.0032702222880389956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,3072,0.006945778098371293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,256,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,128,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,2560,0.009689778089523315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,1024,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,64,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,128,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,32,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,65536,0.050042665667004056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,64,0.00498933345079422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,51200,0.039410667286978826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,1536,0.008719111482302347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,512,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,16384,0.012810666528013019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,12288,0.013501333693663279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,65536,0.1764204502105713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,256,0.005457777943876054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,51200,0.1355982224146525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,16384,0.045718222856521606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,8192,0.011078221930397881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,7168,0.010718222293588849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,128,0.0052782222628593445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,6144,0.009396444592210982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,7168,0.0225582222143809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,32,0.004964444372389051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,6144,0.01995377739270528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,4096,0.00793688909875022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,10240,0.013136000268989138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,3584,0.007293333609898885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,4096,0.01517688896920946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,3072,0.0064195555945237475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,512,0.0035457776652442086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,8192,0.024877332978778418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,3072,0.012774222426944308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,2048,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,2560,0.011397333608733283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,1536,0.004588444613748127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,5120,0.017639999588330586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,1024,0.0038888889054457345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,12288,0.034894221358829074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,768,0.0035751110149754416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,10240,0.02993333339691162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,1024,0.0075804442167282104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,512,0.00351555562681622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,256,0.0034257777863078644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,5120,0.009027555584907532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,128,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,256,0.005384888913896348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,64,0.002833777831660377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,128,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,32,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,1536,0.008854221966531541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,64,0.00526577772365676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,3584,0.014013333453072442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,32,0.004937777916590373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,768,0.007012444237867991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,65536,0.17208445072174072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,51200,0.1352293358908759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,12288,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,2048,0.010020444790522257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,10240,0.00869066682126787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,65536,0.04841511117087471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,8192,0.008010666403505538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,51200,0.03858133488231235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,7168,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,512,0.005969777703285217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,6144,0.007675555845101674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,7168,0.022333333889643352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,5120,0.011072888970375061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,16384,0.04589244392183092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,4096,0.009053332938088311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,12288,0.0347999996609158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,3584,0.008485332959228093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,10240,0.0295448899269104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,3072,0.007354666789372762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,16384,0.010429333481523724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,2560,0.006992888947327931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,3072,0.01276622215906779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,2048,0.005144888742102517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,2560,0.011459555890825061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,1536,0.004588444613748127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,2048,0.010040000081062317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,1024,0.003945777813593547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,1536,0.00871288859181934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,768,0.0038648889296584656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,3584,0.014115555418862237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,8192,0.024832000335057575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,512,0.0033253334048721525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,768,0.0069395556218094295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,6144,0.0199617776605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,512,0.005969777703285217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,128,0.0028231110837724474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,2560,0.005921777751710679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,4096,0.015180443723996481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,64,0.002512000087234709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,32,0.0028400001012616684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,128,0.0052373334765434265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,1024,0.0074728892909155945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,32,0.00508355556262864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,256,0.0031964443624019623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,5120,0.017626666360431247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,256,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,64,0.00508088908261723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,16384,0.7996835708618164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,16384,2.530475616455078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,16384,2.860150231255425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,51200,3.0278568267822266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,12288,1.9398123423258464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,16384,2.4873138003879123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,12288,2.0496533711751304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,12288,0.6549226442972819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,10240,1.6019769244723852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,10240,0.5018026563856337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,12288,1.851668463812934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,10240,1.7201609081692164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,51200,7.8415273030598955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,8192,1.2894923951890733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,8192,0.4222666687435574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,10240,1.4937511020236547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,8192,1.3120862113104927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,7168,1.2729884253607857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,7168,0.3672195540534125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,6144,0.8897519641452365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,7168,1.1082969241672092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,7168,1.06295288933648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,6144,0.3241617679595947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,5120,0.715552912818061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,6144,0.9492710961235894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,8192,1.1977404488457573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,5120,0.2786880069308811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,5120,0.822720898522271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,4096,0.6690115398830838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,51200,11.78505113389757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,5120,0.7150310940212674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,4096,0.23489599757724336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,4096,0.6275795300801595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,3584,0.6660070949130589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,4096,0.5852870941162109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,3072,0.4280542267693414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,3584,0.5581182373894585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,3072,0.1835155619515313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,3072,0.5319004588656956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,2560,0.4197244379255507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,6144,0.8928062650892469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,3584,0.20891466405656603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,2560,0.4054480128818088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,2048,0.3114124404059516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,2560,0.37691556082831484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,3584,0.519670221540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,2048,0.33111998769972056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,1536,0.2448044353061252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,3072,0.4452720218234592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,1536,0.11032711135016547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,1024,0.17033421993255615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,2560,0.1607475545671251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,1536,0.25147467189364964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,1024,0.08812889125612046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,1024,0.18318577607472739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,2048,0.13481067286597356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,768,0.15931555959913465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,512,0.09646844201617771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,1536,0.26336712307400173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,768,0.150672886106703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,512,0.06853244702021281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,512,0.1375057829750909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,768,0.1359457837210761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,512,0.122871994972229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,256,0.06202310985989041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,2048,0.319435543484158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,256,0.09751644399431016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,128,0.11452799373202854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,128,0.07599644528494941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,1024,0.1979964441723294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,128,0.05938844548331367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,32,0.04982044299443563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,256,0.12195555369059245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,768,0.07616266939375135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,128,0.057710223727756076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,32,0.11691021919250488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,51200,8.462022569444445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,64,0.11628177430894639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,64,0.04904799991183811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,256,0.07347555292977227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,65536,2.7829244401719837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,51200,2.2836435106065536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,16384,2.3395582834879556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,51200,6.069911956787109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,65536,7.869370354546441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,16384,0.6434151331583658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,16384,2.332257800632053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,51200,9.089648776584202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,16384,1.8984266916910808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,12288,1.586154619852702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,12288,0.46985334820217556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,12288,1.616471078660753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,65536,12.005031161838106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,10240,1.2462497287326388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,10240,0.40613600942823624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,12288,1.4273120032416449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,10240,1.34625244140625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,8192,1.0910861757066515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,51200,6.474784851074219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,8192,0.3253786563873291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,8192,1.038062201605903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,10240,1.1086621814303927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,7168,0.9293884701199002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,65536,8.381654527452257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,8192,0.9402151107788086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,6144,0.6426080067952474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,7168,0.9200382232666016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,6144,0.2663404411739773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,6144,0.7733626895480685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,7168,0.8256817923651801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,7168,0.2895439995659722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,5120,0.21681777636210123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,5120,0.6431440247429742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,4096,0.4793039957682292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,6144,0.6898026466369629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,4096,0.1828871170679728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,3584,0.38294490178426105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,5120,0.5610133277045356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,3584,0.16393510500590006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,3584,0.45268090565999347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,5120,0.5970044665866429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,3584,0.41511199209425187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,3072,0.14334489239586726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,3072,0.38178488943311906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,3072,0.35327643818325466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,2560,0.12609422206878662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,2560,0.3279333379533556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,4096,0.5203990936279297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,2048,0.25871022542317706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,2560,0.2964666684468587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,4096,0.45424355400933164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,2048,0.10920888847774929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,1536,0.1891484393013848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,2048,0.2612266540527344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,1024,0.13243644767337376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,1536,0.08793777889675564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,2560,0.30960000885857475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,1536,0.19275111622280547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,768,0.1162346601486206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,1024,0.06969866487714979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,1024,0.14876622623867458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,512,0.0757733318540785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,768,0.1258702278137207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,768,0.11987111303541396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,3072,0.33808088302612305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,256,0.052969776921802096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,512,0.055361777544021606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,1536,0.2075359953774346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,512,0.10839200019836426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,1024,0.15609155760871038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,256,0.050435556305779346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,256,0.07815821965535481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,256,0.09665599796507095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,64,0.03865777783923679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,128,0.06146755483415392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,32,0.039340443081325956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,2048,0.25474845038519967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,64,0.09228710995780097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,512,0.09910133149888779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,128,0.04611555404133267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,32,0.09272799889246623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,128,0.04822311136457655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,65536,0.928251584370931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,128,0.09096711211734348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,51200,2.112107594807943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,65536,2.7707653045654297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,51200,0.7019893328348795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,768,0.061503999763064914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,65536,3.583097669813368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,16384,0.7047235700819227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,16384,0.2164693276087443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,12288,0.47055912017822266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,16384,0.6773982048034668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,12288,0.161152892642551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,16384,0.5845262209574381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,12288,0.43706756167941624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,10240,0.36465154753790957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,51200,3.0136462317572703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,10240,0.13990044593811035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,10240,0.4224577744801839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,8192,0.2790195677015516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,10240,0.36055554283989805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,8192,0.1126284466849433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,7168,0.273693323135376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,12288,0.5024195777045356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,8192,0.2975715531243218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,7168,0.10364533132976955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,6144,0.2209022177590264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,7168,0.290829340616862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,7168,0.25421423382229275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,6144,0.0880897773636712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,5120,0.20169511106279162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,6144,0.24725066290961373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,6144,0.22721510463290742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,8192,0.32761067814297146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,65536,2.679778628879123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,4096,0.1430035564634535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,5120,0.20937066608005098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,5120,0.18630399968889025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,4096,0.06434933344523112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,4096,0.1656915611690945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,4096,0.15300267272525364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,3584,0.057766218980153404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,3584,0.14586488405863443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,51200,2.0576328701443143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,3584,0.13836799727545843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,2560,0.09175022443135579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,3072,0.05084000031153361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,3072,0.12782400184207493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,3072,0.12139022350311279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,2048,0.08408000071843465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,3584,0.14119822449154323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,2560,0.1100471085972256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,2560,0.10315377844704522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,1536,0.0570639967918396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,2048,0.039340443081325956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,2048,0.09075378047095405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,3072,0.12088889545864528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,1024,0.04182666540145874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,1536,0.032455109887652926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,1536,0.07295733027988009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,768,0.03365600109100342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,1024,0.026705778307384912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,5120,0.07839288976457384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,1024,0.05525066786342197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,512,0.026916444301605225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,768,0.024076445235146418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,768,0.046329776446024575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,512,0.03669333457946777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,256,0.019720888800091214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,512,0.03868444429503547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,2048,0.08905866410997178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,256,0.02032088902261522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,256,0.03310399916436937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,128,0.018560889694425795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,256,0.03057066599527995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,2560,0.045696000258127846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,64,0.015630222029156156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,512,0.02253866692384084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,128,0.02589599953757392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,32,0.016207110550668504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,64,0.03135200010405646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,1536,0.07107466459274292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,1024,0.05515822105937534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,768,0.04425333274735344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,128,0.03145600027508206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,65536,0.7140844662984213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,65536,1.9357626173231337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,128,0.02009600069787767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,51200,1.5829679701063368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,16384,0.5465564197964138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,32,0.031609776947233416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,51200,0.5544551213582357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,51200,2.2321412828233504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,16384,0.5315475463867188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,16384,0.16717867056528726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,12288,0.4415671030680339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,16384,0.4592675632900662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,12288,0.12797599368625218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,65536,2.009782155354818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,12288,0.44768624835544163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,12288,0.33430666393703884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,10240,0.11307377285427517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,65536,2.8556861877441406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,8192,0.2836719883812798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,10240,0.31634044647216797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,10240,0.2772826618618435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,51200,1.5816533830430772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,8192,0.0910124447610643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,8192,0.25126489003499347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,8192,0.22968710793389213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,7168,0.22049421734280059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,7168,0.19793600506252715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,10240,0.34740887747870547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,5120,0.1605466604232788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,6144,0.17681688732571074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,5120,0.16039110554589167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,7168,0.2290257877773709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,5120,0.1471031109491984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,6144,0.20458666483561197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,4096,0.1288879977332221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,6144,0.07134844197167291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,3584,0.1211893293592665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,5120,0.06059822109010485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,4096,0.13332266277737087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,7168,0.08638311094707912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,3584,0.04494933287302653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,3584,0.10770666599273682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,6144,0.19031288888719347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,4096,0.1220968829260932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,2560,0.08195822106467353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,3072,0.09949866930643718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,3072,0.09533333116107517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,3584,0.11339467101626927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,2048,0.06286933024724324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,2560,0.08289777570300631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,3072,0.10148800081676906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,4096,0.050823999775780566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,1536,0.05156177613470289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,2048,0.07004711363050672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,3072,0.03993155558904012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,1536,0.024813334147135418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,1024,0.03613600134849548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,1536,0.05760000149408976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,1024,0.02107288936773936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,1024,0.04292710953288608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,768,0.02903111113442315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,1024,0.04484000139766269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,2048,0.03105866577890184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,512,0.021640888518757288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,768,0.019111111760139465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,768,0.03468266791767544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,768,0.03746044304635789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,1536,0.056213332547081836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,512,0.028592887851927016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,2560,0.08601778083377415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,512,0.03193777799606323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,128,0.014075555735164218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,256,0.025501334004931982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,256,0.024992000725534227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,2048,0.07043733199437459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,64,0.01239822225438224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,128,0.024695111645592585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,128,0.02111022174358368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,64,0.02418666746881273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,32,0.024414221445719402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,2560,0.03537955549028184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,256,0.016128000285890367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,128,0.015929778416951496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,32,0.012999110751681857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,256,0.016734222571055096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,65536,0.5971608691745335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,65536,1.6935368643866644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,51200,1.2350568771362305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,51200,0.4638853073120117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,65536,2.3393688201904297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,51200,1.7016977734035914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,16384,0.1439102225833469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,16384,0.4488986863030328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,65536,1.6571635140313044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,16384,0.37579022513495547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,16384,0.43816532029045957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,12288,0.12088355753156875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,10240,0.2761911021338569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,12288,0.3143697844611274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,12288,0.27835022078620064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,512,0.01755555636352963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,10240,0.09295466873380874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,8192,0.23183822631835938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,10240,0.2676577832963732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,10240,0.23957599533928764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,8192,0.07521155807707045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,8192,0.21043999989827475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,8192,0.18988800048828125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,6144,0.14608977900611028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,7168,0.18786577383677164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,51200,1.296669324239095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,7168,0.17126222451527914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,5120,0.14633156193627253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,6144,0.1611457798216078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,6144,0.1465724441740248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,5120,0.051015999582078725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,5120,0.1362933317820231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,4096,0.09475377533170913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,5120,0.12659377521938747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,3584,0.09945955541398789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,4096,0.10878311263190375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,7168,0.06700088580449422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,4096,0.10373955302768284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,3072,0.07293955485026042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,3584,0.03797600004408095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,3584,0.09593333138359918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,3584,0.09221422009997898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,3072,0.033850666549470686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,3072,0.08495822217729355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,3072,0.08391200171576606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,7168,0.20090666082170275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,2048,0.053617778751585216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,12288,0.28177955415513783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,2560,0.07147022088368733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,2048,0.05996088849173652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,2048,0.025880000657505457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,1536,0.04113511244455973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,2048,0.06008444229761759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,2560,0.07256622446907891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,1536,0.021406221720907424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,1536,0.04969244533114963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,2560,0.07268799675835504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,2560,0.030709332889980737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,1024,0.018587556150224473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,768,0.023288889063729182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,6144,0.05893777476416694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,768,0.030368000268936157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,768,0.03273866573969523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,768,0.016559110747443307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,4096,0.04261689053641426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,512,0.02507466740078396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,1024,0.03881422347492642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,512,0.015423999892340766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,1536,0.04827644427617391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,1024,0.03186666634347703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,128,0.011742221812407175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,256,0.014469333820872836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,1024,0.03690133492151896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,128,0.02131822208563487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,64,0.010536889235178629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,256,0.013782222237851886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,128,0.018975110517607797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,64,0.020969778299331665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,256,0.022032888399230108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,512,0.018205333087179396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,32,0.02136444383197361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,256,0.02104355560408698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,128,0.01420088940196567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,512,0.027121777335802715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,32,0.01035644445154402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,65536,1.348688867357042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,51200,0.9817804760403104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,51200,0.4019164509243435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,16384,0.3345039950476752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,65536,1.84935908847385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,16384,0.11695644590589736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,65536,0.4996933407253689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,12288,0.26194310188293457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,51200,1.329194704691569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,16384,0.3074391153123644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,51200,1.0114675097995334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,10240,0.20348178015814888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,12288,0.2295537789662679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,10240,0.22560177909003365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,10240,0.07607466644710965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,10240,0.19565599494510225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,16384,0.3608408768971761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,8192,0.17496089140574136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,8192,0.062282668219672314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,65536,1.3162515428331163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,8192,0.15426933765411377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,6144,0.12552444140116373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,12288,0.08988889058430989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,7168,0.152983996603224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,7168,0.1370897822909885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,5120,0.10113510820600723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,6144,0.13271110587649876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,6144,0.12098933590783013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,7168,0.13344000445471868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,5120,0.04271644353866577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,4096,0.07746311028798421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,5120,0.10199644168217976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,7168,0.05546311206287808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,12288,0.2585528956519233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,4096,0.09045688973532783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,4096,0.08489688899781968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,8192,0.18035377396477592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,3584,0.07959377765655518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,3072,0.06964533196555243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,3584,0.07743644714355469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,5120,0.11106666591432358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,3072,0.028304888142479792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,3072,0.06987200180689494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,3072,0.06695466571384005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,2560,0.02586666742960612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,2560,0.06117510795593262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,2048,0.043691555658976235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,6144,0.05020622081226773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,3584,0.03070666723781162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,2048,0.050648887952168785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,1536,0.03368533319897122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,2048,0.04943644338183933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,2560,0.052961776653925575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,1536,0.04095288780000474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,1536,0.04129066732194688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,3584,0.07723999685711332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,2560,0.05894311269124349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,1024,0.031439112292395696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,768,0.019884443945354886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,1024,0.03258311086230808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,2048,0.02221155497762892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,768,0.014879999889267815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,768,0.02765688962406582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,1536,0.019321777754359774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,1024,0.024077332682079736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,512,0.02126311096880171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,512,0.013549333645237817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,4096,0.03551022211710612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,1024,0.01593866613176134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,256,0.01850400037235684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,128,0.010344888601038191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,256,0.013076444466908773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,768,0.025567111041810777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,128,0.012776888906955719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,64,0.009746666583749983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,128,0.017255999975734286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,32,0.010132444401582083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,512,0.023802666200531855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,64,0.01790488925245073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,32,0.017792888813548617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,256,0.019380443625979953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,512,0.014916444818178812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,128,0.01792888840039571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,256,0.012267555627557965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,65536,1.2180382410685222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,51200,0.9280906253390842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,51200,0.34489067395528156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,65536,1.6229306326972113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,51200,1.2403288947211373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,16384,0.10884000195397271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,16384,0.3143208821614583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,16384,0.29023289680480957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,65536,0.4342408974965413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,12288,0.0808746682273017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,10240,0.22053955660925972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,12288,0.2194275591108534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,10240,0.07062844435373943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,10240,0.19038311640421549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,10240,0.18283822801378039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,65536,1.2775760226779513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,51200,0.9572711520724826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,8192,0.05635022454791599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,8192,0.14896889527638754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,7168,0.13696889082590738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,7168,0.1321822272406684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,7168,0.05026488833957248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,12288,0.26745422681172687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,7168,0.15238400300343832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,6144,0.11840000417497422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,6144,0.04444977641105652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,5120,0.09879289070765178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,6144,0.1210835509830051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,5120,0.0381004446082645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,5120,0.09896533356772529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,16384,0.3189191023508708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,6144,0.1157955593532986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,4096,0.032297776805029974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,3584,0.0739368862575955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,4096,0.08233155806859334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,3584,0.07140444384680854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,8192,0.15328178140852186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,3584,0.0737982193628947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,12288,0.22661956151326498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,4096,0.08486311303244697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,3584,0.029440889755884807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,5120,0.09715288877487183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,3072,0.027106665902667578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,3072,0.06299110915925768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,4096,0.08019644684261747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,3072,0.06708444489373101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,2048,0.03776355584462484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,2560,0.054619554016325206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,2560,0.05714044306013319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,2048,0.020852444900406733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,2048,0.045828445090187915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,2048,0.048087110122044883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,3072,0.060468441910213895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,1536,0.01760088900725047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,1536,0.03695022066434225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,2560,0.04988177617390951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,1536,0.04088622331619263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,768,0.017664889494578045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,1024,0.014847111370828418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,1024,0.028350220786200628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,8192,0.1558577749464247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,512,0.01419911119672987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,768,0.013910222384664746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,1536,0.03012622065014309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,512,0.018921777606010437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,256,0.01074044406414032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,1024,0.02144444485505422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,256,0.01684177749686771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,256,0.017636444833543565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,256,0.012079999678664737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,768,0.023378666904237535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,768,0.0266951107316547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,128,0.0163102216190762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,64,0.008408889174461365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,512,0.022699554761250813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,32,0.008923555413881937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,64,0.016195555528004963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,32,0.016248888439602323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,2560,0.024263999528355066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,128,0.009331555830107795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,512,0.01293955577744378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,128,0.01591199967596266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,65536,0.9125750859578451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,1024,0.031297776434156634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,65536,1.3359075122409398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,128,0.012089778151777057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,51200,0.3057333363427056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,51200,1.0264275868733723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,16384,0.28423023223876953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,16384,0.09567733605702718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,16384,0.3159102333916558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,12288,0.20597244633568657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,65536,1.0944186316596138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,12288,0.1981875499089559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,12288,0.0724453330039978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,12288,0.19492444727155897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,10240,0.1703235573238797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,51200,0.6790853076510959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,10240,0.16826044188605416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,8192,0.13803554905785456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,10240,0.1620417833328247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,7168,0.10666310787200928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,8192,0.1352213356229994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,65536,0.38129509819878477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,8192,0.13199822107950845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,6144,0.09788444307115342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,7168,0.0446320010556115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,7168,0.1174773375193278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,51200,0.802495108710395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,6144,0.10366044441858928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,16384,0.256801790661282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,5120,0.08532177739673191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,6144,0.10297244124942356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,4096,0.06687021917766996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,8192,0.051186667548285596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,5120,0.03412444392840067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,5120,0.0873111089070638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,5120,0.08730844656626384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,3584,0.05908444192674425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,4096,0.029230223761664495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,4096,0.07095822360780504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,10240,0.06330400043063693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,3072,0.05164444446563721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,3584,0.026644445127911035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,6144,0.0390897790590922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,3584,0.06577955351935492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,3072,0.05514666769239637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,3072,0.05860444572236803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,2560,0.048080000612470836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,2560,0.022210665875011023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,2048,0.03345155715942383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,7168,0.11897422207726373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,4096,0.07342933283911811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,3584,0.06314311424891154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,2048,0.01945066617594825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,2048,0.04025422202216254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,2048,0.04332888788647122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,1024,0.0199617776605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,1536,0.016878222425778706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,1536,0.03252799974547492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,1536,0.03691555394066705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,768,0.01629599928855896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,1024,0.014164444473054675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,1024,0.02880000074704488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,2560,0.05070399906900194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,512,0.012496000362767113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,768,0.02068800065252516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,768,0.023800000548362732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,512,0.016984888248973422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,2560,0.04369599951638115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,1024,0.025107555919223364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,768,0.013210666676362356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,256,0.017094223035706412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,128,0.010750222537252637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,512,0.012334222594896952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,256,0.012146666646003723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,128,0.0148053334818946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,128,0.011742221812407175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,128,0.015509333875444202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,32,0.010049777726332346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,64,0.014875556031862894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,1536,0.027480888697836135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,256,0.011562666959232755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,512,0.020310221446885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,64,0.009837333526876237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,256,0.015266666809717814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,65536,1.0033138063218858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,65536,0.3394453260633681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,32,0.014493332968817817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,65536,1.0926969316270616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,51200,0.26541156238979763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,16384,0.2191822263929579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,3072,0.024318221542570326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,65536,0.8680177794562446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,51200,0.8206435309516059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,12288,0.17798133691151938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,16384,0.21817066934373644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,12288,0.06212977568308512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,12288,0.17140977912478975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,51200,0.7163377867804633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,12288,0.17608622709910074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,51200,0.6070995330810547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,10240,0.14495200581020778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,10240,0.14281955030229357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,16384,0.2272853321499295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,8192,0.11714222696092393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,7168,0.09648178021113078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,10240,0.13698755370246038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,8192,0.1118417713377211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,7168,0.10361244281133015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,6144,0.08292355802324083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,7168,0.038669334517584905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,16384,0.08123466703626844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,6144,0.08940088748931885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,8192,0.04365866714053684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,10240,0.05933244360817803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,5120,0.07543022102779813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,7168,0.10352089007695515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,5120,0.03018666638268365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,4096,0.0543431109852261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,5120,0.07621155844794379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,3584,0.04744799931844076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,4096,0.025905778010686237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,4096,0.06496977806091309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,3584,0.05526489019393921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,3584,0.023980443676312763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,3072,0.03919644488228692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,3584,0.05722755855984158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,3072,0.021743999587164983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,3072,0.049442665444480054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,2560,0.033997333712048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,3072,0.05127288897832235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,2560,0.01994044416480594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,2560,0.04306044512324863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,2048,0.026918222506841023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,2560,0.044328000810411244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,2048,0.017369776964187622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,2048,0.0358515547381507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,1536,0.022286223040686712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,1536,0.029288887977600098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,1536,0.03157777918709649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,1024,0.015966221690177917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,4096,0.061549332406785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,1024,0.02202044427394867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,1024,0.02516711089346144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,2048,0.0380657778845893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,1024,0.012861332959598966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,768,0.018241778016090393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,768,0.02174844510025448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,1536,0.014632888966136508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,6144,0.08886666430367364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,512,0.015280889140235053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,512,0.018234666850831773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,768,0.013056889176368713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,256,0.01370844410525428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,6144,0.034388442834218345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,8192,0.11418044567108154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,128,0.007337777978844113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,256,0.01073333372672399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,256,0.01482222146458096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,128,0.01311911145846049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,64,0.006664888726340399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,64,0.01311377767059538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,32,0.006663110935025745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,32,0.013150222599506378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,128,0.010709332923094431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,128,0.014705777168273926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,65536,0.576824876997206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,5120,0.06698133548100789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,65536,0.2883226606580946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,65536,0.8812026447719998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,51200,0.5185191366407607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,65536,0.7346800168355306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,16384,0.17912711037529838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,51200,0.2346035639444987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,512,0.010416000253624385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,51200,0.6909333335028754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,16384,0.06967822048399183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,16384,0.19339466094970703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,256,0.008052444292439355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,12288,0.1412924395667182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,10240,0.11706666151682536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,12288,0.13804888725280762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,768,0.012079999678664737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,10240,0.04575377702713013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,51200,0.5795137617323134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,10240,0.11711733871036106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,8192,0.09420622057384914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,512,0.011408000356621213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,8192,0.03671911027696397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,7168,0.07472533649868436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,8192,0.09697600205739339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,8192,0.09661244683795506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,6144,0.06844711303710938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,7168,0.08629333310657078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,7168,0.08789866500430638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,10240,0.11924889352586533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,12288,0.1308622227774726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,5120,0.058240003055996366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,6144,0.07599733273188274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,5120,0.02602311140961117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,5120,0.06324177980422974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,12288,0.05318311187956068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,5120,0.06504978073967828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,7168,0.03308355477121141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,4096,0.051741331815719604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,4096,0.02232444451914893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,3584,0.04047022263209025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,6144,0.030198223061031763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,3584,0.020728000336223178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,3584,0.04935377836227417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,4096,0.04699288805325826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,3072,0.04078933265474107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,3072,0.04418489005830553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,3072,0.018917333748605516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,2560,0.029594666428036157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,16384,0.19094933403862846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,2560,0.017039111918873258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,3072,0.034719109535217285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,2560,0.03856977820396423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,2048,0.030287110143237646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,4096,0.05436533358361986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,1536,0.019665777683258057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,6144,0.07466222180260552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,3584,0.04622133241759407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,1536,0.024344889654053584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,2048,0.02440711193614536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,1536,0.02753066685464647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,1024,0.011887110769748688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,1024,0.01869422197341919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,768,0.011824888487656912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,1024,0.021663111117151048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,2048,0.032871110571755305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,512,0.009684444301658207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,768,0.01107022249036365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,2048,0.015219555960761176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,768,0.018957333432303537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,256,0.007333333293596904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,512,0.010462221999963125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,1536,0.013667555318938361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,512,0.016206221448050607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,128,0.0069297779765393995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,256,0.012120889292822944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,256,0.01310222182008955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,128,0.010019555687904358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,64,0.007424888511498769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,768,0.015896888242827523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,32,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,512,0.013222222526868185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,2560,0.0359191099802653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,1024,0.015096889601813423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,32,0.01145600030819575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,65536,0.5762560102674696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,128,0.011493333511882357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,65536,0.2644960085550944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,128,0.012503999802801343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,65536,0.8060817718505859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,51200,0.458728896247016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,64,0.011468444433477191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,16384,0.17374666531880698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,256,0.010035555395815108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,51200,0.6125049061245388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,16384,0.06648000081380208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,12288,0.11679466565450032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,51200,0.5518337885538737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,16384,0.18074933687845865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,10240,0.11103555228975083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,12288,0.0512586666478051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,12288,0.12873600588904485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,12288,0.13719111018710667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,10240,0.10872888565063477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,51200,0.2129617797003852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,65536,0.7235911157396105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,7168,0.0780328909556071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,8192,0.03520444366667006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,8192,0.0972177783648173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,16384,0.17229066954718697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,6144,0.06779111093944974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,7168,0.03258133265707228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,7168,0.08456710974375407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,10240,0.04399999976158142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,6144,0.0680933329794142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,5120,0.054168889919916786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,6144,0.07564710908465914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,10240,0.11636532677544488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,4096,0.04324355390336779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,5120,0.025472889343897503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,5120,0.05775466892454359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,5120,0.06370221906238131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,3584,0.03867822223239475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,4096,0.04810577630996704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,7168,0.07827822367350261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,4096,0.0544088880221049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,3072,0.034599112139807806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,3584,0.020055111911561754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,8192,0.08781511253780788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,3584,0.04831644561555651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,6144,0.02833333280351427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,2560,0.029618667231665716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,3072,0.018004443910386827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,8192,0.08800444338056777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,2560,0.016374222106403775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,2048,0.025158221522967022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,2560,0.03751733236842685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,4096,0.021749332547187805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,2048,0.028054220808876887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,1536,0.019925332731670804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,2048,0.0322728885544671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,1536,0.022495999932289124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,1536,0.02680444386270311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,1536,0.01332266628742218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,1024,0.01535377734237247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,1024,0.017616889543003507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,2048,0.014827556080288358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,1024,0.02129777769247691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,768,0.012122666670216454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,1024,0.011633777783976661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,768,0.010932444698280759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,768,0.014909333652920194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,512,0.009701333112186855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,768,0.01847466660870446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,3072,0.04279822111129761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,256,0.007696000238259633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,512,0.010267555713653564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,512,0.012481778032249875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,512,0.01573155489232805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,256,0.009769777456919352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,256,0.01107911103301578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,256,0.013166222307417126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,3584,0.04296711087226868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,64,0.006601777755551868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,32,0.007752888732486301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,128,0.012340444657537671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,3072,0.03880177934964498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,64,0.010724444356229572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,32,0.010665777656767102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,65536,0.43425777223375106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,128,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,65536,0.22310845057169595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,51200,0.36168445481194395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,65536,0.6646586524115669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,128,0.010737777584128909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,65536,0.5665884547763401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,51200,0.17725511391957602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,51200,0.5213795767890083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,16384,0.15218844678666857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,2560,0.03345333205329047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,12288,0.10132000181410049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,16384,0.14680533938937718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,10240,0.09373333056767781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,12288,0.04585777719815572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,12288,0.11531821886698405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,12288,0.11200800206926133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,128,0.009845332966910468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,10240,0.0978391104274326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,10240,0.09534755680296157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,8192,0.07866399817996554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,16384,0.06402399804857042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,7168,0.07064000103208753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,8192,0.07902310954199897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,7168,0.028935998678207397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,7168,0.06971289051903619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,6144,0.055536889367633395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,7168,0.07010755274030897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,10240,0.03979377614127265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,16384,0.13267821735805935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,5120,0.047147555483712085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,6144,0.060467554463280566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,6144,0.061780446105533175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,4096,0.036574221319622464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,5120,0.05225244495603773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,51200,0.4414666758643256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,5120,0.05352977911631266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,3584,0.03239733311865065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,4096,0.01958666741847992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,4096,0.04584622383117676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,3584,0.03879911038610671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,3072,0.027454222242037456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,3584,0.018156444032986958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,8192,0.031488001346588135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,3072,0.03427555494838291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,2560,0.024018666810459558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,3072,0.016418667303191293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,4096,0.0429351098007626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,2560,0.029616889026429918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,2560,0.01463733282354143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,2048,0.020004444652133517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,6144,0.02630133264594608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,2048,0.024847999215126038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,2048,0.013455111119482251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,2048,0.02732533382044898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,3072,0.03696533375316196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,1536,0.020395555429988436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,1536,0.012034667034943899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,1024,0.011327111058764987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,2560,0.031947556469175555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,1024,0.016193777322769165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,3584,0.04076355695724487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,768,0.009001777403884465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,1536,0.01462933255566491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,5120,0.0233271105421914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,512,0.008024888734022776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,768,0.010046222143703038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,768,0.0162444445821974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,1536,0.02288177775012122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,512,0.011513777905040316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,512,0.009376000199053023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,1024,0.01829155617290073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,256,0.01039644413524204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,768,0.01370044466522005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,256,0.009006222089131674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,128,0.005919111271699269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,128,0.009721777505344814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,128,0.012312889099121094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,128,0.009027555584907532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,32,0.005763555566469829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,64,0.009754666851626502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,32,0.009752000371615091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,256,0.012477333347002665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,65536,0.4094506634606256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,8192,0.07597688833872478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,64,0.005633777628342311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,65536,0.21095732847849527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,1024,0.01051111105415556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,51200,0.37953511873881024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,65536,0.5681537522210015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,51200,0.16424089007907444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,16384,0.12482310665978326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,51200,0.4502737787034776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,16384,0.06085866689682007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,16384,0.13220355245802137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,16384,0.14493155479431152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,512,0.013786666923099093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,10240,0.07145244545406766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,65536,0.5600853496127659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,12288,0.11071644226710002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,10240,0.036659555302725896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,8192,0.06067555480533176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,12288,0.08390577634175618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,51200,0.4447671042548285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,256,0.00666844430896971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,8192,0.028918223248587713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,7168,0.049455109569761485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,12288,0.04159199860360887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,8192,0.07771022452248467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,7168,0.02659911082850562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,7168,0.06158577733569675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,10240,0.08419377936257257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,7168,0.06932889090643989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,6144,0.05351110961702135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,6144,0.06029778056674533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,8192,0.06894844770431519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,5120,0.04565422071350945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,12288,0.10305600033866034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,4096,0.03031555480427212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,5120,0.052101333936055504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,6144,0.04283110962973701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,3584,0.025245333711306255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,4096,0.03779022230042352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,4096,0.044044444958368935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,10240,0.09479377667109172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,5120,0.03611022233963013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,3584,0.016931555337376065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,3584,0.04007911019855075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,3072,0.03017066584693061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,3072,0.01570666664176517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,2560,0.01850844422976176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,3072,0.035805334647496544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,5120,0.021127111381954614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,2048,0.015599111715952555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,2560,0.014296000202496847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,4096,0.018066666192478605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,2560,0.031382222970326744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,1536,0.012531555361217923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,2048,0.013131555583741931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,3072,0.022981334063741896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,2048,0.027048889133665297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,1024,0.009918221996890174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,1536,0.011620444556077322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,1536,0.01825066738658481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,1536,0.022338666849666174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,768,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,1024,0.010390222072601318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,1024,0.01828799976242913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,2560,0.026421333352724712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,512,0.007337777978844113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,2048,0.02237777743074629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,6144,0.024149333437283833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,3584,0.03408799899948968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,256,0.006317333214812809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,512,0.009364444348547194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,512,0.01351199961370892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,1024,0.014296000202496847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,128,0.0052915554907586836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,256,0.011431111229790581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,768,0.012404444317022959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,768,0.015827554795477126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,128,0.009383999639087254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,64,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,128,0.008999110923873054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,512,0.010493333141009012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,64,0.00902844468752543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,32,0.00903911143541336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,256,0.00941244430012173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,256,0.009002666506502364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,768,0.009764444496896531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,65536,0.16763199700249565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,65536,0.4821768866644965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,32,0.005246222019195557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,51200,0.238663117090861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,16384,0.08014577627182007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,51200,0.12916799386342367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,51200,0.37211378415425617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,65536,0.38664889335632324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,51200,0.3254737854003906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,16384,0.04314133193757799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,12288,0.06849244568083021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,16384,0.10954311158921982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,10240,0.057874666319953076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,12288,0.03341599967744615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,128,0.011057777537239922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,12288,0.08492177724838257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,65536,0.4129662248823378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,10240,0.0728293326165941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,10240,0.07249866591559516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,8192,0.02350666622320811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,8192,0.058984888924492725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,7168,0.04190399911668566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,7168,0.05221244361665514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,12288,0.0853875544336107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,7168,0.05423288875155979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,7168,0.021632888250880774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,10240,0.03040177623430888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,5120,0.030227555168999568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,6144,0.045704887972937695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,6144,0.04758755697144402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,16384,0.11135821872287327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,8192,0.06037155787150065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,5120,0.017472000585661996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,5120,0.04146755403942532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,4096,0.015268445014953613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,4096,0.03257511059443156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,3584,0.0225582222143809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,4096,0.035050666994518705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,8192,0.04727733466360304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,3072,0.019796444310082328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,6144,0.01962577799956004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,3584,0.031848000155554876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,3072,0.025750221477614507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,2560,0.017342221405771043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,6144,0.0336773329310947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,3584,0.014304889572991265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,3584,0.029277334610621136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,2048,0.014743111199802823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,2560,0.024858666790856257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,3072,0.013251555462678274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,3072,0.0284586681260003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,2048,0.01908533275127411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,1536,0.01219555570019616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,2048,0.021691555778185528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,2560,0.022426666484938726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,1536,0.01609244445959727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,1536,0.018586667047606576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,1024,0.012548444171746572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,1024,0.015446222490734525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,4096,0.024217777782016333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,768,0.008076444268226624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,2048,0.010856000085671743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,768,0.011101333631409539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,1536,0.010051555931568146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,1024,0.009575111170609793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,512,0.009491555392742157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,2560,0.012160888976520963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,512,0.012087999946541257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,512,0.008456889126035903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,1024,0.009359999663299983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,256,0.008202667037645975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,256,0.008676444490750631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,128,0.00526577772365676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,256,0.01074044406414032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,512,0.0069439998931354946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,64,0.004629333400064045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,128,0.008375999828179678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,128,0.010718222293588849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,64,0.00831911133395301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,256,0.005340444544951121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,5120,0.03927200039227804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,32,0.0083155557513237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,65536,0.24028621779547799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,51200,0.19592000378502739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,768,0.009015999734401703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,768,0.013795555465751223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,65536,0.3940693272484674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,128,0.008286221987671321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,16384,0.07894755734337701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,65536,0.4113208982679579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,51200,0.12173599667019314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,12288,0.05563644568125407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,51200,0.30845867262946236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,16384,0.09967999988132054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,51200,0.32151465945773655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,16384,0.10873688591851129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,12288,0.03218044506178962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,12288,0.07443910837173462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,12288,0.08408355712890625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,65536,0.15381956100463867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,8192,0.036758220858044095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,10240,0.02906755606333415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,10240,0.07155733638339572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,32,0.004895110925038655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,7168,0.03319733341534933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,16384,0.04196888870663113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,8192,0.05974666939841377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,10240,0.05008800162209404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,7168,0.02092444399992625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,7168,0.0454657773176829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,7168,0.05423644516203138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,6144,0.018939556346999276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,6144,0.04019555449485779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,5120,0.0248364441924625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,6144,0.04703288939264086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,5120,0.016846223009957206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,5120,0.034435556994544134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,4096,0.020696888367335003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,5120,0.041561779048707746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,6144,0.028414222929212783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,4096,0.02867377797762553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,3584,0.018239999810854595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,4096,0.034381333324644305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,10240,0.06306133005354139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,3584,0.014186667071448432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,3072,0.01625866691271464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,3584,0.030784000953038532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,8192,0.051388445827696055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,3072,0.01313688854376475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,3072,0.02294755478700002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,3072,0.0277137772904502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,2048,0.011794666449228922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,2560,0.01204711116022534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,2560,0.020118221640586853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,2560,0.024225778049892847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,1536,0.009465778039561378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,2048,0.01725955473052131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,2048,0.021219554874632094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,3584,0.025810667210155066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,1024,0.008016889293988546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,1536,0.010056888891590966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,1536,0.018239110708236694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,2560,0.014270222849316068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,768,0.00664533343580034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,1024,0.011536888778209686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,1024,0.01481599940194024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,4096,0.01482399966981676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,512,0.005899555567238066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,768,0.008831111093362173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,2048,0.0109120003051228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,768,0.01350311107105679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,256,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,512,0.008314666648705801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,512,0.00868444475862715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,1536,0.014471999473041959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,512,0.012063110868136087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,128,0.004559999952713649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,256,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,256,0.010418666733635796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,128,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,128,0.010043555663691627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,32,0.0042364444169733255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,768,0.010096888575288985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,64,0.007370666497283512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,65536,0.19814666112263998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,32,0.00738844441043006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,8192,0.02295555505487654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,65536,0.29817599720425075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,51200,0.15112622578938803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,51200,0.23476621839735243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,51200,0.20595199531979033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,51200,0.08627822001775105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,16384,0.051479998562071055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,65536,0.10880355702506171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,1024,0.009351111120647853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,12288,0.03711555401484171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,12288,0.05928888585832384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,256,0.007900444169839224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,12288,0.05645955271191067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,64,0.004214222232500712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,16384,0.02939288814862569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,10240,0.04999022351370918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,10240,0.04881777697139316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,8192,0.02528800070285797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,128,0.007292444507280986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,8192,0.04122844338417053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,16384,0.07246222098668416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,7168,0.02236000034544203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,8192,0.040977779361936785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,7168,0.03679822219742669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,7168,0.015321777926550971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,7168,0.036650667587916054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,6144,0.019894222418467205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,12288,0.02400622268517812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,16384,0.07830044296052721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,6144,0.032264888286590576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,5120,0.01758666667673323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,6144,0.03245866629812453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,5120,0.027832888894610937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,5120,0.028806222809685603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,4096,0.014286221729384528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,65536,0.27008620897928876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,4096,0.023355555203225877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,6144,0.014146667387750415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,8192,0.016716443830066256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,3584,0.013424889081054263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,4096,0.011396444506115384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,4096,0.0240639994541804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,3072,0.011887110769748688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,3584,0.011008888483047485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,10240,0.021205332544114854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,3584,0.02264533274703556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,2560,0.010108444425794814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,3072,0.018999111321237352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,3072,0.019880889190567862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,10240,0.030564443932639226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,2048,0.008662222160233391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,2560,0.016689777374267578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,2560,0.018205333087179396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,2048,0.014488000008794995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,1536,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,3584,0.021029333273569744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,3072,0.010459555519951714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,1024,0.006265777680608962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,1536,0.013767110804716745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,2560,0.010021333065297868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,1024,0.009910222556855943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,768,0.005525333185990651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,1024,0.012054222325483957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,2048,0.009409777820110321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,2048,0.015896888242827523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,512,0.004953777624501122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,768,0.008028444316652086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,768,0.01074044406414032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,512,0.007618666523032718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,512,0.007673777639865875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,512,0.010081777969996134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,1024,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,256,0.00904266701804267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,128,0.003915555361244413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,768,0.008731555607583787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,5120,0.012837332983811697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,128,0.00665866666369968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,64,0.0035662220584021676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,128,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,128,0.009000000026490953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,32,0.0036151111125946045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,64,0.006612444503439798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,256,0.0069439998931354946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,32,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,256,0.007299555672539606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,65536,0.10584444469875759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,65536,0.2585902214050293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,1536,0.012168000141779581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,65536,0.25863999790615505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,16384,0.04298044575585259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,51200,0.08505866924921672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,51200,0.20816711584726968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,51200,0.20373066266377768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,16384,0.02864711152182685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,16384,0.07178488704893324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,51200,0.12568977144029406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,12288,0.051296889781951904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,10240,0.02572088936964671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,12288,0.023247111174795363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,1536,0.008893333375453949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,10240,0.043545779254701406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,8192,0.02064533366097344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,10240,0.04767022199100918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,16384,0.06765688790215386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,8192,0.01642666591538323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,8192,0.03615377677811517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,8192,0.040122667948404946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,12288,0.05555999941296048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,7168,0.03240799903869629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,6144,0.016223111086421542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,7168,0.015180443723996481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,10240,0.021003555920388963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,65536,0.1637599998050266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,5120,0.014472888575659858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,6144,0.028304888142479792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,6144,0.03166844447453817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,7168,0.01908177799648709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,4096,0.012191111014948951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,5120,0.024780443973011438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,5120,0.02813777658674452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,256,0.004211555752489301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,4096,0.011390222443474663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,7168,0.03609244359864129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,12288,0.032799998919169106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,4096,0.024059555596775476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,3072,0.010082667072614035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,3584,0.021519111262427434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,5120,0.01277599980433782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,3072,0.010496888723638324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,2560,0.009002666506502364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,4096,0.020651555723614164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,6144,0.014130666851997375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,3584,0.018872888551818002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,3584,0.01110666659143236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,2560,0.010073777702119615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,2048,0.007684444387753804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,2560,0.017581333716710407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,3072,0.01960711181163788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,2048,0.009476444787449306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,1536,0.00666933341158761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,2048,0.01571555601225959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,1536,0.009003555609120263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,1536,0.010919999745157031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,2560,0.014845333165592618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,1024,0.009011555876996782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,1024,0.011756444142924415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,1024,0.008341333104504479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,2048,0.012812444733248817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,3584,0.011047110789351992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,768,0.007977777885066139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,768,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,512,0.004560889055331548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,1536,0.013741333451535968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,256,0.004609777695602841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,512,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,512,0.009704888694816166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,256,0.006443555570311016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,256,0.007297777467303806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,128,0.003943110919660992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,256,0.009343110852771336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,128,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,1024,0.0058133333093590206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,128,0.0087333329849773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,32,0.0038862224254343244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,512,0.007640889121426477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,64,0.005992888990375731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,32,0.005991111199061076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,768,0.005249777601824866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,3072,0.016954667038387723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,51200,0.0883022215631273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,65536,0.216538667678833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,128,0.007325333025720384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,65536,0.24812534120347765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,768,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,16384,0.03408977720472548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,51200,0.17125956217447916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,51200,0.19558489322662354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,65536,0.1099439991845025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,16384,0.02792533238728841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,16384,0.06842577457427979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,12288,0.043819556633631386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,12288,0.022375111778577168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,10240,0.02100533412562476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,64,0.0038986665507157645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,10240,0.037660443120532565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,8192,0.01757244434621599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,16384,0.0577075547642178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,8192,0.031034668286641438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,8192,0.015909334023793537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,65536,0.10425866974724664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,7168,0.0158906661801868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,7168,0.02791733211941189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,10240,0.020277332928445604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,7168,0.03376444511943393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,6144,0.013906665974193148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,7168,0.014685332775115967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,6144,0.013427555561065674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,6144,0.024711999628278945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,5120,0.012505778008037143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,6144,0.030001779397328694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,8192,0.03820266657405429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,4096,0.010664888554149203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,5120,0.02141511109140184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,5120,0.02642222245534261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,51200,0.08402044243282741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,3584,0.009906666974226633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,4096,0.022263111339675054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,3584,0.016551999582184684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,3584,0.020974222156736586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,12288,0.05352355705367195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,3584,0.010729778144094678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,10240,0.045704887972937695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,3072,0.018823999497625563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,12288,0.0251857770813836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,2560,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,2560,0.013148444394270578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,3072,0.008709333009190029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,2560,0.009735110733244155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,4096,0.01791200041770935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,2048,0.011612444288200803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,3072,0.010260444548394945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,2048,0.009108444054921469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,1536,0.006137777947717243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,2048,0.014845333165592618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,1536,0.009758222434255812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,1536,0.00868622213602066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,1024,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,1024,0.008350222474998897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,1024,0.01144088887506061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,768,0.004755555755562252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,1024,0.007976888782448238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,768,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,5120,0.012453333371215396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,768,0.007687110867765214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,512,0.0042453333735466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,768,0.01038755559258991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,512,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,512,0.009394666386975182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,256,0.0038764443662431505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,2560,0.016916443904240925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,1536,0.013458666702111563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,256,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,256,0.007080000307824876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,4096,0.01110755569405026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,128,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,128,0.00869599978129069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,128,0.007084444165229797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,64,0.0031582222630580268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,32,0.0031884445084465873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,65536,0.06574222114351061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,2048,0.007009777757856581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,64,0.005611555443869696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,32,0.005662222289376789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,51200,0.05288799934917026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,3072,0.014852444330851236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,65536,0.103584885597229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,65536,0.17385244369506836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,16384,0.01902933253182305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,65536,0.24351554446750215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,51200,0.08285066816541883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,51200,0.13632888264126247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,512,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,51200,0.1920320060518053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,16384,0.06731555859247844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,12288,0.03571288784344991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,128,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,10240,0.01645066671901279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,12288,0.05199822121196323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,10240,0.020211555891566806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,10240,0.030572444200515747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,8192,0.013742222554153867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,10240,0.044666667779286705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,16384,0.026759111218982275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,16384,0.04610933197869194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,8192,0.015875554747051664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,8192,0.03708533445994059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,12288,0.02166666587193807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,7168,0.02276800076166789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,6144,0.011381333072980246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,7168,0.03316888875431485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,6144,0.013458666702111563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,6144,0.020282665888468426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,5120,0.009752000371615091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,6144,0.029106666644414265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,12288,0.018199111024538677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,5120,0.012099555797047086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,8192,0.02532977859179179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,5120,0.025790222816997107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,4096,0.011074666347768573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,4096,0.015400888191329109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,3584,0.007699555820888943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,256,0.008692444198661381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,4096,0.021682666407691106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,3072,0.007287110719415877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,3584,0.010710222025712332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,3584,0.014129777749379476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,3584,0.020409777760505676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,2560,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,3072,0.010048888623714447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,4096,0.008359111017651027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,3072,0.01850488947497474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,2048,0.005647111270162795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,2560,0.011742221812407175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,7168,0.014557333456145393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,2048,0.009104889300134447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,2048,0.010394666757848527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,2048,0.014512888259357877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,3072,0.012774222426944308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,1536,0.008799999952316284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,7168,0.012460444536474017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,1024,0.0042462220622433555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,2560,0.009729777773221334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,2560,0.016407999727461074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,1024,0.008026666939258575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,1024,0.011400000088744693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,1536,0.00510044437315729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,768,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,512,0.0035537779331207275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,768,0.010084444450007545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,5120,0.017848889032999676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,1536,0.013117333253224691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,256,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,512,0.0074648890230390764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,1024,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,768,0.0039057777159743835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,256,0.005591111050711737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,256,0.0070062221752272705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,256,0.008295111358165741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,128,0.0069848886794514125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,128,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,64,0.0031822222388452957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,128,0.00815466625822915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,32,0.0031884445084465873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,1536,0.00865955568022198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,512,0.005975111077229182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,65536,0.05300266544024149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,32,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,128,0.003216000066863166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,65536,0.1723404460483127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,65536,0.24316443337334526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,768,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,51200,0.13554044564565024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,51200,0.08335555262035793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,16384,0.013175110850069257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,64,0.0052675555149714155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,51200,0.19154043992360434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,12288,0.014127999544143677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,16384,0.02642755541536543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,16384,0.06691288948059082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,51200,0.043214221795399986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,12288,0.03500888744990031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,12288,0.021581333544519212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,12288,0.051679111189312406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,8192,0.01310222182008955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,10240,0.020052444603708055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,10240,0.029723554849624634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,10240,0.04397777716318766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,7168,0.01184622198343277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,8192,0.025011556016074285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,512,0.009376889301670922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,16384,0.04601777924431694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,6144,0.010392888552612728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,7168,0.014526221487257214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,65536,0.10335999727249146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,10240,0.012715555727481842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,5120,0.009349333743254343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,6144,0.013436444103717804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,6144,0.020285333196322124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,6144,0.028946667909622192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,5120,0.012087999946541257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,5120,0.017631111873520743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,5120,0.025449777642885845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,4096,0.015249777171346875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,3584,0.007289778027269576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,7168,0.022555554906527202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,7168,0.03283111254374186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,3584,0.014157333307796054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,3584,0.010392000277837118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,3072,0.006608888920810487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,3584,0.02032355633046892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,4096,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,2560,0.005984000033802456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,8192,0.015871110889646742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,3072,0.018263111511866253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,4096,0.01108000013563368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,2048,0.005564444594913059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,2560,0.009683555199040307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,2560,0.016164445214801364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,2048,0.00906755526860555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,1536,0.004903111192915174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,3072,0.010133333504199982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,8192,0.036426666710111834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,2560,0.01145155562294854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,1024,0.004007110993067424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,1536,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,1536,0.012806221842765808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,1024,0.008018666671382057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,2048,0.014480888843536377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,1024,0.011054221954610614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,3072,0.012807110945383707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,768,0.0069297779765393995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,512,0.003555555724435382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,768,0.010048000348938836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,4096,0.021523555119832356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,512,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,256,0.003248888792263137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,512,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,768,0.003925333420435588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,256,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,128,0.0029004443850782183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,256,0.008344000412358178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,768,0.007670222057236566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,128,0.005064888960785336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,64,0.0028702221396896574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,128,0.008107555409272512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,512,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,2048,0.010050666828950247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,1024,0.007637333538797166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,65536,0.04829688866933187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,32,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,256,0.0070186663005087115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,16384,0.015282667345470853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,1536,0.00902133352226681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,12288,0.01275466630856196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,51200,0.13525599903530544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,10240,0.011277333729796939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,16384,0.0457857780986362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,8192,0.010016889207892949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,64,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,7168,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,8192,0.02480799953142802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,65536,0.17242844899495444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,6144,0.008880889250172509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,7168,0.022878222995334204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,5120,0.008033778104517195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,32,0.0028799999919202593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,12288,0.034621334738201566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,6144,0.020111110475328233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,3584,0.00700711127784517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,5120,0.017715555098321702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,51200,0.039298666848076716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,4096,0.015053333507643806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,2560,0.005647999958859549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,3584,0.014130666851997375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,3072,0.012773333324326409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,2560,0.011406222151385413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,1536,0.004613333278232151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,2048,0.010033778018421596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,4096,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,1024,0.003921777837806278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,768,0.0038720000949170855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,10240,0.029861334297392104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,512,0.0032177778581778207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,3072,0.006309333360857434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,1024,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,768,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,128,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,512,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,2048,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,256,0.005595555735958948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,32,0.002891555635465516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,65536,0.04793511165512932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,128,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,64,0.004979555391603046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,51200,0.03762044509251913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,32,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,16384,0.010375111467308467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,12288,0.008671110702885522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,1536,0.008744888835483128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,10240,0.008347555167145198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,65536,0.17236088381873238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,51200,0.13530311319563124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,16384,0.04586844311820137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,7168,0.011370666325092316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,10240,0.029496000872717962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,64,0.0026551110463009942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,8192,0.02476444509294298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,5120,0.009369778136412302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,6144,0.020192889703644645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,4096,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,5120,0.017666666044129264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,3584,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,256,0.0032302221904198327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,3072,0.006593777901596493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,12288,0.034494221210479736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,2560,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,6144,0.010425777898894416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,2048,0.005250666704442766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,7168,0.022509333160188463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,2560,0.011415999796655444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,2048,0.010038221875826517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,1024,0.004006222304370669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,1536,0.008685333033402761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,768,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,8192,0.011742221812407175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,4096,0.01516533394654592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,1024,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,3584,0.014097778333557976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,256,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,768,0.006796444455782573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,3072,0.012747555143303342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,128,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,64,0.0029146667155954572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,32,0.0028444443725877335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,64,0.0052151112920708126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,32,0.004961777892377642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,512,0.003544888976547453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,128,0.0029057777590221832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,512,0.005971555494599872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,1536,0.004573333180612988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,128,0.004963555683692296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,16384,0.8006053500705295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,16384,2.3224542405870223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,12288,1.911144044664171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,256,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,12288,0.5791964530944824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,51200,2.787968953450521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,12288,1.8951839870876734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,16384,2.5839191012912326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,10240,1.5193191104465062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,8192,0.40242399109734434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,10240,1.5146071116129558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,8192,1.2932462692260742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,10240,0.4967039955986871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,7168,0.9874809053209094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,7168,0.37241776784261066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,7168,1.0320702658759224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,51200,7.332542419433594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,6144,0.8842488924662272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,5120,0.8663022253248426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,5120,0.27524622281392414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,4096,0.628877321879069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,5120,0.7189120186699761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,6144,0.3124142222934299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,8192,1.1692017449273002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,3584,0.49428266949123806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,3584,0.19847910934024385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,4096,0.5825875600179037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,3584,0.5025422308180068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,3072,0.4899164305792914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,6144,0.9888782501220703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,2560,0.35489688979254824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,4096,0.21781244542863634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,2048,0.29763910505506724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,51200,10.311901516384548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,2048,0.29913777775234646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,1536,0.27479643291897243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,3072,0.1794311073091295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,1024,0.19589332739512125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,3072,0.43993687629699707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,1536,0.23711021741231283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,768,0.14105689525604248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,1024,0.17791288428836396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,2560,0.15261156029171413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,512,0.0999528898133172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,768,0.07243021991517809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,768,0.14310755994584826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,256,0.0721857746442159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,512,0.06621955500708686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,2048,0.12809511025746664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,128,0.05814044343100654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,256,0.06052977508968777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,256,0.11070577965842353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,64,0.053265776899125844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,128,0.10409422053231134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,1024,0.08285422457589044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,64,0.10439111126793756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,2560,0.378473785188463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,32,0.10563288794623481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,512,0.1242613395055135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,1536,0.11688178115420872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,32,0.056440889835357666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,128,0.05772711171044243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,16384,1.8057910071478949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,16384,0.6065955691867405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,51200,2.2498647901746964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,16384,2.0370506710476346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,12288,0.47757509019639754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,65536,2.7590329911973743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,12288,1.4115627076890733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,51200,5.632189432779948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,12288,1.5165369245741103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,10240,1.1911226908365886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,10240,0.3946026696099176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,8192,0.9003040525648328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,65536,7.370679219563802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,8192,0.3249217669169108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,8192,0.9362373352050781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,10240,1.1622230741712782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,7168,0.868716451856825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,6144,0.24511377016703287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,6144,0.7578497992621528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,7168,0.29315111372205943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,5120,0.2230799992879232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,5120,0.5741510921054417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,4096,0.5144649081759983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,4096,0.17331911457909477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,4096,0.47082842720879453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,3584,0.44054489665561253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,51200,7.824195014105903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,3584,0.15937777360280356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,6144,0.6808435651991102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,3584,0.39171377817789715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,7168,0.8001546859741211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,3072,0.14146578311920166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,2560,0.32268444697062176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,3072,0.33881510628594297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,2560,0.12041332986619736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,65536,10.70976087782118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,2560,0.28578933080037433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,2048,0.10226933161417644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,2048,0.23541866408454049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,1024,0.12391911612616645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,1536,0.19069511360592314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,3072,0.38947378264533145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,768,0.0901519987318251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,1024,0.14042399989234075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,512,0.07835377587212457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,768,0.1130524476369222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,2048,0.2473022143046061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,256,0.06167466772927178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,1536,0.18414133124881318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,512,0.0981244444847107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,1536,0.08095466428332858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,128,0.046513779295815356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,256,0.04923466510242886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,256,0.08797955513000488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,128,0.04678044385380215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,32,0.041915555795033775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,1024,0.06472533278995089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,768,0.058669335312313504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,64,0.08286488718456693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,32,0.08370221985710992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,512,0.05375644564628601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,5120,0.5511484675937229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,64,0.040229333771599665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,128,0.08247822523117065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,65536,0.9005875057644315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,16384,0.6325742403666178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,51200,0.6629981994628906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,16384,0.20367556148105195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,65536,2.476292504204644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,51200,2.1038284301757812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,12288,0.6340728865729438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,12288,0.1563297775056627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,51200,2.5667546590169272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,65536,3.297306696573893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,10240,0.4114835527208116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,8192,0.30613689952426487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,8192,0.11094311210844253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,7168,0.28230754534403485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,16384,0.750732421875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,7168,0.09869422515233357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,10240,0.4063093397352431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,7168,0.2724844349755181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,6144,0.08670310841666327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,6144,0.22603911823696562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,5120,0.21646043989393446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,5120,0.07426133420732287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,4096,0.15407555633121067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,5120,0.1900320053100586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,4096,0.06160889069239298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,4096,0.1509608825047811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,8192,0.29855823516845703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,6144,0.21284177568223742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,3584,0.13258311483595106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,12288,0.4527217547098796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,2560,0.10011200110117595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,3072,0.11600889099968804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,10240,0.13382311662038168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,2048,0.07084977626800537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,2560,0.09961777925491333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,1536,0.05834666887919108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,3584,0.14211555322011313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,2048,0.08240533537334867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,3584,0.05529866615931193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,3072,0.12709332836998835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,1536,0.06679822338951959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,768,0.031145777967241075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,1024,0.02470933397610982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,1024,0.0496053331428104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,512,0.023863999380005732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,768,0.022457778453826904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,768,0.04025244381692674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,256,0.019707555572191875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,2048,0.03760177890459696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,1024,0.03867644402715895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,256,0.030018667380015056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,128,0.015799999237060547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,3072,0.04831199844678243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,64,0.013428444663683573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,2560,0.04309244288338555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,32,0.013794666363133324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,512,0.020392888122134738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,64,0.02823822365866767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,512,0.033077332708570696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,32,0.02858755654758877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,1536,0.03017066584693061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,128,0.018805333309703402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,128,0.028636445601781208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,65536,0.6951733165317111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,16384,0.4876702096727159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,51200,0.5460631052652994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,51200,1.4312417772081163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,16384,0.17750310897827148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,65536,2.0217857360839844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,16384,0.5157030953301324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,12288,0.4015519883897569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,12288,0.12223111258612739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,10240,0.3279537889692518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,51200,1.9615796407063801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,10240,0.11418399545881484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,8192,0.26727554533216685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,10240,0.2908239894443088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,8192,0.09082666370603774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,256,0.018757333358128864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,7168,0.07721066474914551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,8192,0.2701386610666911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,7168,0.2020204464594523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,5120,0.15795111656188965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,6144,0.17313955889807808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,5120,0.05893066856596205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,5120,0.1450079944398668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,4096,0.047247111797332764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,7168,0.23193955421447754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,3584,0.12026045057508682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,6144,0.20882488621605766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,3072,0.09884266720877753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,3584,0.10349422030978733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,3072,0.03804533349143134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,3072,0.08966755867004395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,2560,0.08244178030225965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,4096,0.11622399753994411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,2560,0.0336408880021837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,3584,0.04273688793182373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,6144,0.0675519969728258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,2048,0.029072887367672388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,1536,0.0513608886135949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,65536,2.2989253997802734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,1536,0.02408444384733836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,1024,0.03353066576851739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,2560,0.07684533463584052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,2048,0.06713955269919501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,768,0.026964444253179762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,2048,0.06501955456203885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,768,0.03173955612712436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,512,0.020245333512624104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,768,0.018575110369258456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,1536,0.051523556311925255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,512,0.01680444512102339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,512,0.02623822291692098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,128,0.012431110772821637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,256,0.015505777464972602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,1024,0.020242666204770405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,64,0.011021333436171213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,128,0.015596444408098856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,32,0.01127377814716763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,12288,0.353200011783176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,64,0.02238311039076911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,32,0.022335110439194575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,256,0.01588088936275906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,1024,0.038943999343448214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,128,0.022673777408070032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,4096,0.10743288861380683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,65536,1.5473608440823023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,16384,0.36209776666429305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,51200,1.3707653681437175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,51200,0.4612728754679362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,16384,0.13607911268870035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,12288,0.35693955421447754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,65536,2.13025877210829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,65536,0.5633431010776097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,51200,1.4256079991658528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,12288,0.10498755507998997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,12288,0.2862933211856418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,8192,0.20132354895273843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,10240,0.2924151155683729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,10240,0.2513546678755018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,7168,0.17812889152102998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,8192,0.19246221913231742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,7168,0.0656026668018765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,6144,0.14996355109744602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,7168,0.17293243938022188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,16384,0.4086337884267171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,6144,0.05746755335066053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,6144,0.14651555485195583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,5120,0.04845688740412394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,5120,0.12425866391923691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,10240,0.08897955550087823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,3584,0.08719555536905925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,8192,0.07458133167690702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,3584,0.03582933213975694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,3072,0.08520355489518906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,3584,0.08828800254397923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,256,0.02349511120054457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,2560,0.06597155332565308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,3072,0.03243466549449497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,5120,0.12208799521128337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,2048,0.054377777708901294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,2560,0.029715554581748113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,4096,0.11482400364345974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,2560,0.06682222419314914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,2048,0.05477866530418396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,1536,0.021524444222450256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,1024,0.028516444895002577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,1536,0.04433688852522108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,4096,0.09927022457122803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,768,0.02274666726589203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,1024,0.017853332890404593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,512,0.017050666941536795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,1024,0.03445333242416382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,768,0.01630311045381758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,256,0.015680889288584392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,512,0.015095111396577625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,3072,0.0773457752333747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,512,0.0229102224111557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,1536,0.03988177908791436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,2048,0.025272889269722834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,256,0.020364445116784837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,64,0.009558222360081142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,128,0.013912000589900546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,4096,0.040170666244294904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,64,0.019324445062213473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,32,0.01946311030122969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,768,0.027705777022573683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,128,0.011050666371981302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,256,0.014137778017255994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,65536,1.114598274230957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,32,0.009730666875839233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,128,0.01964977714750502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,51200,0.8302728864881727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,16384,0.3036666711171468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,16384,0.3122008906470405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,16384,0.11467733648088242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,65536,1.614705827501085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,51200,1.2473404142591689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,65536,0.47887553109063047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,12288,0.23344977696736655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,8192,0.16778488953908285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,10240,0.0731875565316942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,10240,0.19781600104437935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,8192,0.05921600262324015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,7168,0.15364000532362196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,12288,0.22219555907779268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,8192,0.15734044710795084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,6144,0.10612177848815918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,7168,0.05352977911631266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,10240,0.19203021791246203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,6144,0.046683556503719754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,6144,0.12007644441392686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,5120,0.03933511177698771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,5120,0.10120444165335761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,51200,0.3734133243560791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,3584,0.07298488749398126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,4096,0.033365332418017916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,4096,0.08116800255245633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,3072,0.06297777758704291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,3584,0.02993866801261902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,2560,0.05068622363938225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,3072,0.027079999446868896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,3584,0.07295644283294678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,12288,0.08660622437795003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,2048,0.04052977760632833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,2560,0.024530667397711012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,1536,0.03122488988770379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,2048,0.021838222940762837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,2048,0.04601600103908115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,1536,0.018908444378111098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,1536,0.03736089004410638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,1024,0.015821332732836407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,1024,0.029038220643997192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,768,0.018922666708628338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,7168,0.13998044861687556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,768,0.01440533333354526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,768,0.023385778069496155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,4096,0.08690133359697129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,256,0.014535110857751636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,512,0.013397333522637686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,512,0.01945600079165565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,1024,0.024461333950360615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,128,0.009032888544930352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,256,0.012552888856993781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,256,0.017264889346228707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,128,0.012466666599114736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,32,0.00830844458606508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,128,0.01687111126052009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,5120,0.11044355233510335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,512,0.014787554740905762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,32,0.016588444511095684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,2560,0.05492800143029955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,65536,0.9531306160820855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,64,0.008222222328186035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,51200,0.9332631429036459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,16384,0.28901423348320854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,51200,0.33124176661173504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,51200,1.0025866826375325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,16384,0.10081688563028972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,12288,0.21084888776143393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,16384,0.4030382103390164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,12288,0.0900720026757982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,10240,0.17902933226691353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,12288,0.2113493283589681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,10240,0.06807111369238959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,8192,0.14361155033111572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,10240,0.22835199038187662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,3072,0.06386933061811659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,7168,0.12123910586039226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,8192,0.141792893409729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,6144,0.11467911137474908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,7168,0.048010667165120445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,65536,1.457273801167806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,5120,0.09966933064990574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,6144,0.04595999916394552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,6144,0.10781421926286484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,4096,0.07544177770614624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,65536,0.42078577147589785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,3584,0.07102221912807889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,4096,0.07327999671300252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,64,0.016516443755891588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,3072,0.05613599883185493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,3584,0.028350220786200628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,3072,0.025507556067572698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,2560,0.04661155409283108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,5120,0.03696622120009528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,5120,0.09104533327950372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,2048,0.03923555546336704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,2560,0.0233324451578988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,8192,0.05501155720816719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,2048,0.041165331999460854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,3584,0.06473510795169406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,7168,0.1237128840552436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,3072,0.05667022201750013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,1024,0.021720000439220007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,1536,0.017412443955739338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,4096,0.03136000037193298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,1024,0.025918222135967676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,2048,0.02015999952952067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,1536,0.029077334536446467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,512,0.013785777820481194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,768,0.021332444416152105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,512,0.012576888832781049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,1536,0.033896889951494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,256,0.010086221827401055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,512,0.017937777770890128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,128,0.00872622181971868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,256,0.011908444265524546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,1024,0.014736000034544202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,768,0.017229333519935608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,64,0.008990222381220924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,128,0.015509333875444202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,32,0.00903911143541336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,768,0.013735999663670858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,64,0.015178667174445258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,32,0.015164444843928019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,65536,0.36811643176608616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,128,0.011935110721323224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,51200,0.8829004499647352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,16384,0.24766400125291613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,51200,0.28951822386847603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,51200,0.9100862079196506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,12288,0.17791555987464058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,16384,0.2406648794809977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,12288,0.07052266597747803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,2560,0.048897776338789195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,65536,1.1693556043836806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,10240,0.05980622106128269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,8192,0.13455643918779162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,10240,0.15265511141883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,65536,1.1133484310574002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,7168,0.10473422209421794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,8192,0.12353955374823676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,7168,0.04328533344798618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,6144,0.09265600310431586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,7168,0.10869155327479045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,10240,0.15971910953521729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,5120,0.08350577619340684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,6144,0.09427822298473781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,5120,0.033257779147889875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,4096,0.06383555465274386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,8192,0.04852977726194593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,4096,0.0640595555305481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,16384,0.09790400001737808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,3584,0.025905778010686237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,12288,0.18341867129007974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,3072,0.05156800150871277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,256,0.015544000599119397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,2560,0.04014577799373203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,3072,0.0236151119073232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,3072,0.050380445188946195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,2560,0.021299555897712708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,2560,0.043390221065945096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,4096,0.02850488821665446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,1536,0.0265600002474255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,2048,0.03675288955370585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,3584,0.05698844459321764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,6144,0.03832711113823785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,1024,0.018213333355055917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,1024,0.013958222336239286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,5120,0.07942755354775323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,768,0.014647111296653748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,1024,0.022854222191704646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,2048,0.03404444456100464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,768,0.013131555583741931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,768,0.019256888164414298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,512,0.012119111087587146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,256,0.009001777403884465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,1536,0.016399110356966656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,1536,0.030001779397328694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,256,0.014472888575659858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,128,0.011424889167149862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,64,0.007266666326257918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,512,0.01183377785815133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,32,0.007346666521496243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,3584,0.058964444531334766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,64,0.013801777528391944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,32,0.013805333111021252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,512,0.0158595558669832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,256,0.011391110718250275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,65536,0.3228000005086263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,128,0.014035556051466199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,51200,0.6461342175801595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,16384,0.18821956051720512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,2048,0.018823999497625563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,51200,0.25758398903740776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,51200,0.7297564612494575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,16384,0.07730399900012545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,16384,0.2400533358256022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,128,0.008299555215570662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,12288,0.1566444502936469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,10240,0.05270666546291775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,65536,0.8760835329691569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,10240,0.1340684493382772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,65536,0.95431825849745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,8192,0.10531466537051731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,12288,0.15966044531928167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,10240,0.12868800428178576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,7168,0.03753777676158481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,6144,0.07325510846243964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,8192,0.09506489170922174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,6144,0.034254223108291626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,6144,0.08158222171995375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,8192,0.0420648886097802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,4096,0.05308177736070421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,12288,0.06531377633412679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,4096,0.05570755402247111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,3584,0.04582488867971632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,5120,0.05684622128804525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,3584,0.02329866588115692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,3072,0.03875911235809326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,5120,0.06856977939605713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,2560,0.033703111939960055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,3072,0.04428622126579285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,7168,0.0928879976272583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,2560,0.019002666076024372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,2048,0.027106665902667578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,2048,0.016689777374267578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,1536,0.021354667014545862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,3584,0.04974933465321859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,7168,0.08703733152813381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,1536,0.014668444792429606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,1024,0.01587733295228746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,3072,0.021088000800874498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,1024,0.012698666916953193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,768,0.012875555290116204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,1024,0.020192889703644645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,2560,0.038576887713538274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,768,0.012100444071822695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,768,0.016954667038387723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,512,0.01130311108297772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,2048,0.03229599859979417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,5120,0.02940000096956889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,256,0.010709332923094431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,256,0.012773333324326409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,1536,0.026718222432666357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,64,0.006424888968467712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,128,0.010618666807810465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,128,0.012426666915416718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,32,0.006439111298984951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,4096,0.02574488851759169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,64,0.012305777933862476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,32,0.012437333663304647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,256,0.00889511075284746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,512,0.014329777823554145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,65536,0.6444400151570638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,128,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,51200,0.3837306764390733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,16384,0.16278133127424452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,51200,0.22300799687703451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,16384,0.0662355555428399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,16384,0.17293422751956514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,512,0.009912889036867354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,12288,0.05207822057935926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,12288,0.12832178009880915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,10240,0.04637866550021701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,10240,0.11042666435241699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,65536,0.28622666994730633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,7168,0.07375199927224053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,8192,0.036172442966037326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,8192,0.08899111217922634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,12288,0.12365067005157471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,51200,0.5877502229478624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,10240,0.10251555840174358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,6144,0.03033866816096836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,6144,0.06769955820507474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,8192,0.07998133367962308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,4096,0.039502223332722984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,5120,0.05802310837639702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,4096,0.021938666701316833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,7168,0.03308355477121141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,3584,0.0355751117070516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,7168,0.07832355631722344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,3072,0.031847112708621554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,3584,0.020113777783181932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,5120,0.052101333936055504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,3584,0.042576889197031655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,2560,0.026928888426886663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,3072,0.018216888109842937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,2048,0.021807110971874658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,3072,0.037787556648254395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,2560,0.03289955523278978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,1536,0.01758311192194621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,2048,0.0280693338976966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,4096,0.047691557142469615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,1024,0.013757333159446716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,1536,0.01346666696998808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,1536,0.02280355493227641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,768,0.011446221835083432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,1024,0.011747555600272285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,1024,0.017647999856207106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,512,0.009194666312800514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,768,0.011077333655622272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,768,0.014642667439248828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,5120,0.02545333405335744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,256,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,512,0.010069333016872406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,2048,0.014918223023414612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,128,0.006229333165619109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,6144,0.062048892180124916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,256,0.01107022249036365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,64,0.00535733335547977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,128,0.00980177770058314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,32,0.005588444570700328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,64,0.010775110787815519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,32,0.01074044406414032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,512,0.012615111139085559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,65536,0.7764568858676486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,65536,0.5874640146891276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,65536,0.2578675482008192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,51200,0.43062133259243435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,65536,0.6838186581929525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,16384,0.14366043938530818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,51200,0.20432444413503012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,12288,0.10521777470906575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,51200,0.6146213213602701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,16384,0.15539111031426325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,10240,0.09803199768066406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,256,0.009696000152164036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,12288,0.1164826684527927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,8192,0.06950310866038005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,10240,0.0440577765305837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,7168,0.06608177555931939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,8192,0.03437155485153198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,8192,0.08006399869918823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,7168,0.031383110417260066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,2560,0.016586666305859882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,7168,0.0707519981596205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,16384,0.06346578068203397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,5120,0.04736266533533732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,6144,0.02797422144148085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,6144,0.061568001906077065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,5120,0.02462933295302921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,5120,0.05246488915549385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,4096,0.021367111139827307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,128,0.011052444577217102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,3584,0.019648000597953796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,6144,0.054005334774653115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,3072,0.03421866562631395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,3072,0.017678222722477384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,3584,0.03198488884501987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,10240,0.0989875528547499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,2560,0.016347555650605094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,2560,0.029884444342719182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,2048,0.025056888659795124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,2048,0.014601776997248331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,2560,0.023435556226306494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,4096,0.042768889003329806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,1536,0.012792000340090858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,1536,0.020790222618314955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,1024,0.011861333416567909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,1024,0.01106844428512785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,768,0.01038044442733129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,1536,0.016351110405392118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,768,0.013740444348918067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,512,0.007975999679830339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,768,0.010863111250930362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,3584,0.038189331690470375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,512,0.010060444474220276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,512,0.011697777443461947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,1024,0.016193777322769165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,128,0.005991111199061076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,128,0.01015377789735794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,128,0.009416888985368941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,2048,0.01997244523631202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,64,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,32,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,256,0.009383999639087254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,64,0.010039110978444418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,256,0.01073155552148819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,32,0.010064000056849586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,65536,0.5276453230116103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,51200,0.40512977706061465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,3072,0.028809779220157202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,65536,0.6008426878187392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,16384,0.11895822154151069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,51200,0.17214666472540963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,12288,0.0564897788895501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,12288,0.09010310967763265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,16384,0.06006311045752632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,10240,0.08787644571728176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,12288,0.04435288906097412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,12288,0.1034266683790419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,65536,0.21479645040300158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,10240,0.03892533315552606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,10240,0.08753422233793472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,8192,0.031089779403474595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,7168,0.05825422207514445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,8192,0.0717724429236518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,4096,0.03830400109291077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,7168,0.06377510892020331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,51200,0.4867439799838596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,6144,0.025864000121752422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,6144,0.05570133527119955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,5120,0.04141955574353536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,8192,0.06581155459086101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,4096,0.03223022156291538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,5120,0.022328888376553852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,256,0.00666933341158761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,3584,0.028740445772806805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,4096,0.019554666346973844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,4096,0.03967377874586318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,3072,0.02390577726893955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,16384,0.13939021693335638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,3072,0.015831111205948722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,3072,0.03192799952295091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,2560,0.021783111823929682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,2560,0.014524444937705994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,2048,0.01866666641500261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,2560,0.027466666367318895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,2048,0.013410666750537025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,1536,0.014711999230914645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,3584,0.03532888823085361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,1536,0.011884444289737277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,1536,0.019333332777023315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,1024,0.010219555762079027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,768,0.008647110727098253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,1024,0.010375111467308467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,5120,0.04729511009322273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,1024,0.014967999524540372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,512,0.007295110987292395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,768,0.010029333333174387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,768,0.012715555727481842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,256,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,512,0.009377777576446533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,3584,0.01703288820054796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,128,0.005274666680230035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,256,0.009018667042255402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,64,0.005087999833954705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,256,0.009679999616410997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,128,0.00868711123863856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,32,0.00489688871635331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,128,0.009383999639087254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,64,0.009352000223265754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,32,0.009359999663299983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,65536,0.4074666765001085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,6144,0.04627466532919142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,51200,0.3187662230597602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,2048,0.022834666901164587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,512,0.011031111081441244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,65536,0.5059768888685439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,51200,0.41897066434224445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,16384,0.12060177326202393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,12288,0.09213155508041382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,10240,0.06484088632795545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,7168,0.029616889026429918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,16384,0.10681155655119155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,10240,0.07741333378685845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,8192,0.05667199691136678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,51200,0.15780533684624568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,7168,0.04587999979654948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,8192,0.029530667596393164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,8192,0.06257510847515531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,6144,0.03948355714480082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,12288,0.040582223070992365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,7168,0.055656890074412026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,5120,0.0329075555006663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,65536,0.20309244261847603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,6144,0.04842933350139194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,4096,0.025778666138648987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,10240,0.03590933481852213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,12288,0.08147288693322076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,16384,0.05271644393603007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,3584,0.024296000599861145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,4096,0.03446311089727614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,3072,0.021000888612535264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,3584,0.03108533223470052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,3072,0.01516266663869222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,3072,0.027229333917299908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,2560,0.018910222583346896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,2560,0.0141004439857271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,2048,0.01588088936275906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,4096,0.017882666654056974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,3584,0.01642400026321411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,1536,0.012998222476906247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,6144,0.02382755610677931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,1536,0.017089777522616915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,1024,0.010031110710567897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,1536,0.011412444214026133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,5120,0.020614221692085266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,2560,0.02370133333735996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,1024,0.010255110760529837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,1024,0.013520888984203339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,512,0.007300444775157505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,768,0.009732444253232744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,2048,0.020455110404226515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,7168,0.025962667332755193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,512,0.009080000221729279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,512,0.010027555955780877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,128,0.005120888766315249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,256,0.00868088917599784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,256,0.008904000123341879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,128,0.00867377801073922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,128,0.008722667064931657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,2048,0.01290844463639789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,768,0.011399110986126794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,32,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,32,0.00869688888390859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,65536,0.3433600001864963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,64,0.004883555488453971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,65536,0.15745155016581217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,51200,0.21951999929216173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,65536,0.44705867767333984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,64,0.00868088917599784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,256,0.00591822216908137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,51200,0.12376266055636936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,51200,0.35502666897243923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,12288,0.061669336424933545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,16384,0.1039395531018575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,10240,0.05128888951407539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,768,0.008838222258620793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,10240,0.029430223835839167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,8192,0.0432515541712443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,16384,0.08297599686516656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,8192,0.023601777023739282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,8192,0.054874665207333036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,16384,0.041745778587129384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,6144,0.030685332086351182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,7168,0.02147377696302202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,12288,0.0328453348742591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,7168,0.04936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,5120,0.02644355595111847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,6144,0.01906222270594703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,4096,0.022184888521830242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,5120,0.017069333129458956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,10240,0.06765244404474895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,3584,0.01844888925552368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,4096,0.031256887647840716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,7168,0.035800000031789146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,3072,0.016705777910020616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,3584,0.013983999689420065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,3072,0.012832000023788877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,6144,0.0432933337158627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,2560,0.014901333385043673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,5120,0.03715377714898851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,4096,0.015185778339703878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,2560,0.021684444612926904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,3584,0.027905778752432928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,2048,0.01829333272245195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,1536,0.010482666393121084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,1536,0.010055999788973067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,5120,0.041393776734670006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,1536,0.015503111812803479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,1024,0.009350222018029954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,2048,0.0122417774465349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,1024,0.01237066669596566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,768,0.009038222332795462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,512,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,12288,0.07941333452860515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,512,0.008364444805516137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,256,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,2560,0.011980444192886353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,256,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,128,0.004576888763242298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,768,0.01039644413524204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,128,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,64,0.003941333128346337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,1024,0.008303110798199972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,32,0.004231111043029361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,64,0.007962666451931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,32,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,65536,0.22724710570441353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,256,0.008387555678685507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,65536,0.14840711487664118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,51200,0.20944800641801622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,65536,0.33886223369174534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,3072,0.0248364441924625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,16384,0.06750488943523832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,51200,0.12473155392540826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,12288,0.05160355567932129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,51200,0.2697128984663221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,16384,0.08769422107272679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,10240,0.04228622383541531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,12288,0.03176622258292304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,2048,0.010823110739390055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,8192,0.03164088726043701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,10240,0.05520977907710605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,128,0.008008889026112026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,512,0.009301332963837517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,7168,0.027543110979927912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,8192,0.022279111875428095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,8192,0.04552533229192098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,6144,0.024705777565638225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,7168,0.02068088948726654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,768,0.007122666471534305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,5120,0.021469333105617102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,7168,0.0403164459599389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,4096,0.017843556072976854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,6144,0.03558311197492812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,5120,0.016336000627941556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,3584,0.015915556086434256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,5120,0.030697776211632624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,4096,0.014726221561431885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,10240,0.028526223368114893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,4096,0.025516443782382544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,3584,0.013628444737858243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,3584,0.022829333941141765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,2560,0.012472888661755456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,3072,0.012759110993809171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,3072,0.02039377722475264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,16384,0.038600001070234514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,2048,0.010616888602574667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,6144,0.018340445227093168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,2560,0.011772444678677453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,2048,0.010648888846238455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,2048,0.015555555621782938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,12288,0.06487644380993314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,1024,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,1536,0.010141332944234213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,1536,0.01314399970902337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,768,0.006665777828958299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,1024,0.009140444298585257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,512,0.005611555443869696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,768,0.008723555339707268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,768,0.009111111362775167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,256,0.004566222015354368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,512,0.008395555946562026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,512,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,128,0.004190222256713443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,256,0.008014221986134848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,256,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,64,0.0038951109680864546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,128,0.008011555506123437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,32,0.0038986665507157645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,3072,0.01423111061255137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,64,0.0070017774899800616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,65536,0.1905519962310791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,32,0.007391999993059371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,51200,0.15017244550916883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,1536,0.009006222089131674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,65536,0.2846435440911187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,16384,0.05222577850023905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,51200,0.22169600592719185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,12288,0.040756444136301674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,128,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,16384,0.07277688715193006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,12288,0.02295733326011234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,12288,0.05549600058131748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,65536,0.10087022516462539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,10240,0.02109866672092014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,8192,0.027671110298898485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,2560,0.01794399983353085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,51200,0.08176622125837538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,7168,0.02475555572244856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,8192,0.03913155529234145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,7168,0.015501333607567681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,6144,0.021881777379247878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,7168,0.03481600019666884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,5120,0.01812799937195248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,1024,0.010484444598356882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,6144,0.030689779255125258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,4096,0.01566488875283135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,5120,0.02647288971477085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,10240,0.04732444551255968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,3584,0.013803555733627744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,4096,0.011741333537631564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,8192,0.01754755609565311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,3072,0.011852444046073489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,3584,0.01071466671095954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,16384,0.027779555983013574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,3584,0.020281778441535104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,2560,0.010760889285140567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,3072,0.010388444695207808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,2048,0.009352000223265754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,6144,0.014614222778214348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,2560,0.009937778115272522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,1536,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,2560,0.016026667422718473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,2048,0.009425777528021071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,1024,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,1536,0.008796444369686974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,1536,0.011763555308183035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,1024,0.008281778130266402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,1024,0.009564444422721863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,768,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,768,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,512,0.005323555734422471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,10240,0.032163557079103254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,256,0.00453955555955569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,512,0.0074648890230390764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,5120,0.01315822203954061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,2048,0.013815110756291283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,256,0.007338666253619724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,256,0.006637333167923822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,64,0.003540444291300244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,768,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,32,0.003879111260175705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,64,0.006640888750553131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,3072,0.017960889471901786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,512,0.007369777394665613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,128,0.003939555750952827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,65536,0.2407155566745334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,51200,0.12299199899037679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,128,0.007296889192528195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,128,0.006468444648716185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,16384,0.04086844457520379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,51200,0.08080444733301799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,51200,0.18935378392537436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,65536,0.13932711548275417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,12288,0.02847466535038418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,16384,0.0625200006696913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,10240,0.02420266634888119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,12288,0.0479964448346032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,65536,0.09957600302166408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,4096,0.022008889251285132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,8192,0.019686222076416016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,8192,0.016237333416938782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,7168,0.018024888303544786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,32,0.006594666590293248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,16384,0.02725511127048069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,6144,0.016315556234783597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,7168,0.015505777464972602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,12288,0.022760889596409265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,5120,0.013931555880440606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,6144,0.013808888693650564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,6144,0.026769777139027912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,5120,0.012475555141766867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,5120,0.023051554958025616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,4096,0.01146755533085929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,8192,0.033765332566367254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,4096,0.019495111372735765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,3584,0.010747555229398938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,3072,0.009604444106419882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,3072,0.010404444403118556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,3072,0.016121778223249648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,2560,0.00867022242810991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,10240,0.04094666573736403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,2048,0.0075502221783002215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,2560,0.009704888694816166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,2560,0.014144000079896716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,3584,0.011012444065676795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,2048,0.012445333103338877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,1536,0.008664888640244802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,1536,0.010668444136778513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,1024,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,10240,0.02052533295419481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,4096,0.012077333198653327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,1024,0.008729777402347988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,768,0.008012444608741337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,512,0.004274666723277834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,7168,0.030491554074817236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,512,0.0069955554273393415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,256,0.003564444267087512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,512,0.007607111500369177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,1024,0.008006222546100616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,128,0.0035582222044467926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,256,0.006320888797442119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,128,0.007130666739410824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,1536,0.006908444480763541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,3584,0.017667555146747164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,64,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,256,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,65536,0.11512622568342422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,2048,0.009376889301670922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,64,0.003565333369705412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,51200,0.09129422240787083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,65536,0.0989111132091946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,32,0.003539555602603488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,16384,0.0348293317688836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,51200,0.07982310983869764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,51200,0.15945777628156874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,768,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,12288,0.027064000566800434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,16384,0.0532622238000234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,10240,0.023445333043734234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,12288,0.021691555778185528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,768,0.007695111135641734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,12288,0.04082755578888787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,8192,0.01979466610484653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,10240,0.019948444432682462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,7168,0.017619555195172627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,10240,0.03516799873775906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,8192,0.015769778026474845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,8192,0.0292577776643965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,7168,0.014498665928840637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,7168,0.02645955483118693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,5120,0.013853333062595792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,6144,0.023132445083724126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,5120,0.012083555261294046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,5120,0.020096888144810993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,65536,0.20384089152018228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,4096,0.017215111189418368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,6144,0.0158942225906584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,3584,0.010255999863147736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,3584,0.010453333457310995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,32,0.006113777971929974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,3072,0.009651555783218807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,3584,0.01568000018596649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,2560,0.008411555654472774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,3072,0.010068444742096795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,4096,0.011571555501884885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,2048,0.007345777418878343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,2560,0.012601777911186218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,4096,0.01106844428512785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,2048,0.011085333095656501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,2048,0.009053332938088311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,6144,0.013453333742088742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,16384,0.026312889324294195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,1024,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,1536,0.008424888882372115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,1536,0.010040000081062317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,1024,0.008013333711359236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,2560,0.009487999810112847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,1024,0.008042666647169325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,512,0.004443555656406614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,768,0.007714666426181793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,768,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,256,0.0035377778112888336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,512,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,1536,0.006699555450015598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,128,0.003540444291300244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,256,0.0070062221752272705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,64,0.003516444315512975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,128,0.007035555938879649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,128,0.005974222388532426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,32,0.0032275555034478507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,128,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,65536,0.06807021962271796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,64,0.00555644432703654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,32,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,65536,0.09851288795471191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,512,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,16384,0.019472888774342008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,51200,0.07944533560011122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,51200,0.12373422251807319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,256,0.005919999960396025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,12288,0.016247999336984422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,16384,0.02515022291077508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,16384,0.04210844304826525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,3072,0.014069333672523499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,12288,0.02109333376089732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,8192,0.013560888667901358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,10240,0.01960533360640208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,10240,0.028155555327733357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,51200,0.05472266674041748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,8192,0.015511999527613321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,8192,0.023362666368484497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,6144,0.01071999967098236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,7168,0.014259555273585849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,7168,0.021190222766664293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,5120,0.009690666364298927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,6144,0.018944000204404194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,10240,0.018608889646000333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,768,0.004903111192915174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,4096,0.008354666332403818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,12288,0.0326222214433882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,3584,0.007641777396202087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,4096,0.014488889111412896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,3584,0.010391111175219217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,3072,0.00702044450574451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,6144,0.013234666652149625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,3072,0.009777777724795872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,5120,0.012098666694429187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,4096,0.010770666930410596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,2560,0.01072622256146537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,2048,0.005673777725961473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,2560,0.009475555684831407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,3584,0.013262222210566202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,1536,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,2048,0.009707555174827576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,1536,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,1024,0.004229333251714706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,5120,0.01660355594423082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,65536,0.1565671099556817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,1024,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,768,0.00388711111413108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,768,0.007631999750932057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,2048,0.009030222064918941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,768,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,512,0.007156444092591603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,3072,0.012087110843923358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,512,0.005981333139869902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,7168,0.012129777835475074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,256,0.00544799988468488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,1024,0.008010666403505538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,128,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,128,0.005469333380460739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,2560,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,64,0.005120000077618493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,256,0.007089777953094906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,128,0.0032222221295038858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,32,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,65536,0.09866844283209907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,512,0.0035333335399627686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,65536,0.15637511677212185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,256,0.003543111185232798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,32,0.003216000066863166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,51200,0.12298399872250027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,16384,0.024431111084090337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,65536,0.048241777552498706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,16384,0.041660444604025945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,12288,0.02104266650146908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,64,0.0028693332440323303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,12288,0.03265955713060167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,51200,0.07919555240207247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,1536,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,16384,0.01754577789041731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,10240,0.027848889430363972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,10240,0.019654222660594516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,7168,0.010398222340477837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,8192,0.015527110960748462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,12288,0.013501333693663279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,6144,0.009690666364298927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,7168,0.02102488941616482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,51200,0.03963111175431146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,5120,0.009341333475377824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,8192,0.009861333502663506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,5120,0.016543999314308167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,4096,0.007919111185603673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,5120,0.012079111403889127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,8192,0.023028444912698533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,3584,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,4096,0.014505777094099255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,10240,0.011742221812407175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,3584,0.010409778190983666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,6144,0.018805333309703402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,3072,0.009921777579519484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,2560,0.005921777751710679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,3072,0.012089778151777057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,2048,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,2560,0.009359111388524374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,2560,0.010738666686746808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,1536,0.004893333133724001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,2048,0.009667555491129557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,1536,0.008649778034951951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,3584,0.01312711089849472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,1024,0.0042239998777707415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,1024,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,768,0.0037760001917680106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,1024,0.007279111279381647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,768,0.007663110891977946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,512,0.003370666669474708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,7168,0.014469333820872836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,512,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,512,0.005913777897755305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,256,0.0052933332820733385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,256,0.006981333096822103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,128,0.0028897778441508612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,768,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,128,0.0069875551594628235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,128,0.005230222311284807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,6144,0.013146667016877068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,32,0.0029057777590221832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,64,0.004954666727119022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,65536,0.04368533359633552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,32,0.005230222311284807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,51200,0.03536355495452881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,2048,0.008712000317043727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,16384,0.013099555340078143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,64,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,3072,0.006609777609507243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,12288,0.011839110818174152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,10240,0.01073155552148819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,12288,0.03205510973930359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,8192,0.010047111246320937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,4096,0.010776888993051318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,7168,0.009679999616410997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,8192,0.023012444376945496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,6144,0.009030222064918941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,51200,0.12263645066155328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,1536,0.00846400029129452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,5120,0.00869599978129069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,7168,0.020977778567208186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,6144,0.018566222654448617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,5120,0.016516443755891588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,10240,0.027506666051016912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,3072,0.006613333192136552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,3584,0.01310488912794325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,2560,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,3072,0.012088889049159156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,2048,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,4096,0.007695111135641734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,1536,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,2048,0.009700444837411245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,1024,0.003899555653333664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,4096,0.014150222142537435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,768,0.0038720000949170855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,65536,0.1569253338707818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,768,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,2560,0.0107351111041175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,256,0.003216000066863166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,512,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,128,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,256,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,64,0.0028568889117903183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,1024,0.007552000383536021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,32,0.0028293333533737394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,16384,0.04131644301944309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,65536,0.040715555349985756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,3584,0.006960888703664144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,51200,0.0336133340994517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,1536,0.008383110993438298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,65536,0.16007999579111734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,128,0.005092444519201915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,16384,0.010379555324713389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,51200,0.12227288881937663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,12288,0.009355555805895064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,64,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,16384,0.041088889042536415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,12288,0.03194577826393975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,10240,0.02739199995994568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,7168,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,8192,0.02307555576165517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,6144,0.010084444450007545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,7168,0.02088711162408193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,5120,0.01036444471942054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,512,0.0032026666320032547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,4096,0.008991999758614434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,5120,0.016545777519543965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,8192,0.007968000239796108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,32,0.005299555758635203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,3072,0.007289778027269576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,3584,0.01310400002532535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,2560,0.006324444380071428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,3072,0.012113778127564324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,2048,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,2560,0.01073333372672399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,6144,0.0188044442070855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,1536,0.004556444370084339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,1024,0.003994666867785984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,2048,0.009707555174827576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,3584,0.008060444560315874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,768,0.003552888830502828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,512,0.003206222214632564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,1536,0.00868000007337994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,4096,0.014256000518798828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,1024,0.007324444750944774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,128,0.002862222285734283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,768,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,64,0.0028293333533737394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,32,0.00254488881263468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,256,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,128,0.005192000004980299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,64,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,10240,0.0086666668454806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,32,0.004931555440028508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,256,0.002895111011134254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,512,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,16384,0.7125973171657987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,16384,2.49688720703125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,16384,2.5346577962239585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,51200,2.6506765153672958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,12288,1.690992037455241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,16384,2.129464043511285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,12288,0.5204328960842556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,10240,1.3249084684583876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,12288,1.6233476003011067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,10240,1.468422253926595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,10240,0.4505217870076497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,51200,7.048414018419053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,8192,1.1079378128051758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,12288,1.8122969733344183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,10240,1.3306222491794162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,8192,1.1167706383599174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,8192,0.36561955346001523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,7168,0.9067573547363281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,7168,0.3361653221978082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,8192,1.0319378111097548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,6144,0.941431999206543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,6144,0.2850737836625841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,6144,0.8569262292650012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,51200,9.940254211425781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,7168,0.9949181874593099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,5120,0.9844524595472547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,6144,0.7801991038852267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,4096,0.5314533445570204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,5120,0.7077671156989204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,7168,0.8690666622585721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,4096,0.5563413302103678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,4096,0.20678755972120497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,3584,0.4323555628458659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,5120,0.6492480172051324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,4096,0.5188195440504286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,3584,0.1816506650712755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,5120,0.249492433336046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,3584,0.4786657757229275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,3072,0.16098754935794407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,3072,0.4331173366970486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,2560,0.3922204441494412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,2560,0.14551822344462076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,2560,0.33047644297281903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,3072,0.3701919979519314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,2048,0.2915982140435113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,3584,0.4474346902635362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,1536,0.19644089539845785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,2048,0.2716248830159505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,2560,0.3540186617109511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,3072,0.4010817739698622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,1536,0.09643999735514323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,1536,0.23167199558681914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,1024,0.14398577478196886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,51200,7.29842546251085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,1536,0.22130578094058564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,1024,0.07583466503355238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,1024,0.17370488908555773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,1024,0.16200889481438532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,2048,0.12032266457875569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,768,0.14017777972751194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,768,0.13455022705925837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,2048,0.31216443909539116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,512,0.12104088730282253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,512,0.06118488974041409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,512,0.1086568898624844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,768,0.12669777870178223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,256,0.0559119979540507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,128,0.05520266625616285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,256,0.08660178052054511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,512,0.09008977810541789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,64,0.04762399858898587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,128,0.053173333406448364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,128,0.10108888811535305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,32,0.04911288950178358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,128,0.0676399999194675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,64,0.1012257801161872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,256,0.07019111182954577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,32,0.10251200199127197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,256,0.10763377613491482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,768,0.06793333424462213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,51200,1.9496248033311632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,65536,2.5958711836073136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,51200,5.364946577284072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,16384,2.0059778425428605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,16384,2.0827609168158636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,65536,6.963547600640191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,16384,0.5490204493204752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,12288,1.315008905198839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,51200,7.5720630221896705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,12288,0.43567644225226504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,12288,1.3661119672987196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,10240,1.114068455166287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,16384,1.64136536916097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,12288,1.2241129345364041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,65536,10.076603359646267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,10240,0.3521920045216878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,10240,1.1542844772338867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,8192,0.849088880750868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,8192,0.28880977630615234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,8192,0.9152498245239258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,10240,1.0014933480156791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,8192,0.7925919956631131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,7168,0.7924008899264865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,6144,0.5933901998731825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,51200,5.583638932969834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,6144,0.2226311100853814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,7168,0.6798826853434244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,5120,0.4802284240722656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,7168,0.26480889320373535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,65536,7.208285013834636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,6144,0.6031529108683268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,5120,0.19278844197591147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,5120,0.5426879988776313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,5120,0.4936800003051758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,3584,0.3317635589175754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,4096,0.16065866417355007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,6144,0.679257763756646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,4096,0.45015642378065324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,4096,0.4100213315751817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,3584,0.14399200015597874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,3584,0.3904693391587999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,4096,0.41380977630615234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,3072,0.127454227871365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,3072,0.33173955811394584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,7168,0.7039733462863498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,3072,0.31592800882127553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,2048,0.21401245064205596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,2560,0.27887733777364093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,2560,0.2616044415367974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,3072,0.28507643275790745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,2048,0.09700089030795628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,1536,0.17297511630588105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,3584,0.34939911630418563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,1536,0.08153688907623291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,1536,0.1826862229241265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,1536,0.17370488908555773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,2560,0.1123466624153985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,768,0.09049333466423883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,1024,0.14222221904330784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,2048,0.23083644443088105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,768,0.11106933487786187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,2560,0.2461893293592665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,512,0.06626666916741265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,1024,0.12001866764492458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,512,0.04970133304595947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,256,0.05020266771316528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,1024,0.06158577733569675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,256,0.0856506692038642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,256,0.06912799676259358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,768,0.10737866825527614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,128,0.03888355692227682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,512,0.09570578071806167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,512,0.08743377526601155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,128,0.04333777891265022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,128,0.08034133248858981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,64,0.03539289037386576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,32,0.037392000357309975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,1024,0.12893866168128118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,2048,0.2231422265370687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,64,0.08046222395367093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,256,0.04538755615552267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,32,0.08118400308820936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,128,0.05491377909978231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,768,0.05482133229573568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,65536,0.7999973297119141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,51200,1.9416959550645616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,65536,2.4719306098090277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,51200,0.6232675446404351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,16384,0.6390542454189724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,65536,2.9425315856933594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,16384,0.22232267591688368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,16384,0.6136435402764214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,51200,2.430639902750651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,12288,0.14656710624694824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,10240,0.36963823106553817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,12288,0.4415031009250217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,12288,0.3866417672899034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,10240,0.1285546620686849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,8192,0.301162666744656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,12288,0.5301555527581109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,10240,0.3313982221815321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,51200,1.8331413269042969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,8192,0.1001360019048055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,8192,0.27003823386298287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,7168,0.25389599800109863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,7168,0.08916266759236653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,65536,2.36826048956977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,6144,0.2048142221238878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,6144,0.08008266819847955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,6144,0.21688266595204672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,16384,0.5152293311225044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,5120,0.18176089392768013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,6144,0.19868355327182344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,4096,0.1540506680806478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,5120,0.06874577866660224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,5120,0.18190311061011422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,5120,0.16763110955556235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,10240,0.36467909812927246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,7168,0.22940444946289062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,3584,0.11507733662923177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,4096,0.14505155881245932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,4096,0.13976977931128606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,3072,0.1003333330154419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,3584,0.05091999967892965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,3584,0.12872178024715847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,8192,0.2875200112660726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,3584,0.12292799684736465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,3072,0.04581422275967068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,3072,0.1135279999838935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,3072,0.10906044642130534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,2560,0.09691288736131455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,2560,0.09391378031836616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,4096,0.05647644731733534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,2048,0.03525777657826742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,2048,0.0806995564036899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,2048,0.0793715582953559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,7168,0.2685973379347059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,1536,0.028441776831944782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,1536,0.0643999973932902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,1536,0.06480089161131117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,2048,0.06281599733564588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,1024,0.048560887575149536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,1024,0.023279110590616863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,1024,0.05061599943372938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,1536,0.052022222015592784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,768,0.03887910975350274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,512,0.027148443791601393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,768,0.04260622130499946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,2560,0.08289511336220635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,1024,0.03945066531499227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,256,0.021418665846188862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,512,0.03602755400869582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,2560,0.040700445572535195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,128,0.01700088878472646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,768,0.03236622280544705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,256,0.02908888790342543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,256,0.028484443823496502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,64,0.014518222875065274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,128,0.017658667431937326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,768,0.021232888102531433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,32,0.014448000325096978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,128,0.02382044494152069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,64,0.027664888236257765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,32,0.02732266651259528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,512,0.0321706665886773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,256,0.01794755624400245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,128,0.027790221903059218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,65536,0.6268311076694065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,512,0.01942933268017239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,65536,1.876231935289171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,51200,1.6137172910902235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,51200,0.48175822363959414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,65536,2.268284479777018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,16384,0.14482577641805014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,51200,1.8115954928927953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,16384,0.4610080189175076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,16384,0.40984445148044163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,12288,0.3382559882269965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,12288,0.116848005188836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,12288,0.30282754368252224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,51200,1.3730381859673395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,10240,0.2889671060774061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,8192,0.19432532787322998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,10240,0.2741813394758436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,10240,0.25431021054585773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,12288,0.38386045561896426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,8192,0.07928533024258085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,8192,0.21893510553571913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,65536,1.838338640001085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,8192,0.20815822813245985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,7168,0.06987200180689494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,7168,0.19360355536142984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,7168,0.1791528860727946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,10240,0.09502577781677246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,6144,0.16694311300913492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,6144,0.061108443472120494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,6144,0.15693067179785833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,5120,0.05289511217011345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,4096,0.1158062219619751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,5120,0.1409484412935045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,5120,0.13334043820699057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,6144,0.16379200087653265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,3584,0.09779200288984512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,4096,0.11257244480980767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,4096,0.10967377821604411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,3072,0.08842933177947998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,3584,0.09950755702124701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,5120,0.13171288702223036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,3584,0.09819910923639934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,2560,0.0730480021900601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,3072,0.08801244364844428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,16384,0.4478942023383246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,4096,0.04487199915779961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,2048,0.05731733640034994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,3584,0.04055288765165541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,2048,0.06241510974036323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,7168,0.2101262278027005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,3072,0.08721511231528388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,1536,0.044144888718922935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,2560,0.07556533151202731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,2560,0.07495555612776014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,1536,0.05030044582155016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,1536,0.02278311053911845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,1536,0.05218311150868734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,1024,0.03827288746833801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,1024,0.01939022209909227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,1024,0.04122311207983229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,3072,0.03597777750757005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,2048,0.02725511127048069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,768,0.03074399961365594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,512,0.017992888887723286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,768,0.03466399841838413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,1024,0.03070577647950914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,256,0.018427555759747822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,512,0.029112001260121662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,768,0.02258133391539256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,256,0.02286400066481696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,128,0.01628800067636702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,256,0.014854222536087036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,768,0.01817244456874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,2560,0.032249778509140015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,128,0.022034666604465906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,128,0.015269334117571512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,32,0.015598222613334656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,512,0.016467556357383728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,512,0.025784000754356384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,64,0.021664000219768945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,32,0.021709332863489788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,256,0.02221777704026964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,2048,0.06553955872853597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,64,0.015517334143320719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,128,0.019310222731696237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,65536,1.3625848558213975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,16384,0.3463475439283583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,51200,0.4172355598873562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,65536,2.074795616997613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,51200,1.4660409291585286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,16384,0.3766133255428738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,16384,0.12816888756222194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,16384,0.338734229405721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,12288,0.2767884466383192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,65536,1.480822245279948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,51200,1.035339567396376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,12288,0.0985635585255093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,51200,1.1134409374660916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,12288,0.24940888086954752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,10240,0.08412710825602214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,10240,0.28188355763753253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,10240,0.21216177940368652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,8192,0.18753333886464438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,7168,0.1551591157913208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,7168,0.16490844885508218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,7168,0.15165332953135172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,7168,0.059453335073259145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,6144,0.1454213327831692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,65536,0.507945802476671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,8192,0.06738577948676215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,8192,0.17345778147379556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,6144,0.05327199896176656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,6144,0.1427688863542345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,5120,0.1226222250196669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,6144,0.13030489285786948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,5120,0.046032889021767505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,5120,0.11870488855573867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,5120,0.1115857760111491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,4096,0.09553422530492146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,3584,0.0885511106914944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,10240,0.2336497836642795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,3584,0.08539644214842056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,4096,0.09317421913146973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,3072,0.07446489069196913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,3584,0.03425511055522495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,3584,0.08314666483137342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,3072,0.0306951105594635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,3072,0.07539021968841553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,3072,0.0739519993464152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,4096,0.0921528869205051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,2048,0.04868533213933309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,2560,0.027729777826203242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,2560,0.06400355365541247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,12288,0.2796497874789768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,1536,0.03797422183884515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,2048,0.023806222611003455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,8192,0.18125243981679282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,2560,0.06291821930143568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,1536,0.04384266667895847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,4096,0.03751466671625773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,1536,0.044736001226637095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,1024,0.017299556069903903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,1024,0.03345333205329047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,768,0.02025422288311852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,1024,0.03535822365019057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,2048,0.05354666709899902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,2048,0.05448177787992689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,768,0.03016622198952569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,1536,0.020599999361568026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,1024,0.025920000341203477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,512,0.022464000516467627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,512,0.014344000154071383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,2560,0.06561422348022461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,256,0.019894222418467205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,128,0.010724444356229572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,256,0.020057777563730877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,768,0.01554844445652432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,512,0.015427556302812366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,64,0.009694221946928237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,128,0.019205333458052743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,32,0.009680888719028896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,128,0.017484444710943434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,256,0.012094222009181976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,512,0.02568799919552273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,32,0.01891644464598762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,256,0.013349333571063148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,768,0.02683199942111969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,128,0.013449778159459433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,65536,1.1918453640407987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,64,0.018959111637539335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,51200,0.8950835333930122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,65536,1.4337484571668837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,51200,0.33773689799838597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,16384,0.3243529001871745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,51200,1.0967120064629448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,16384,0.3128933376736111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,65536,0.44340888659159344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,65536,1.253017743428548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,16384,0.28910401132371694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,12288,0.08301599820454915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,12288,0.23481689559088814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,12288,0.21452800432840982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,10240,0.1920266681247287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,8192,0.16194755501217312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,10240,0.18232177363501656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,16384,0.11480355262756348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,8192,0.05742577711741129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,7168,0.1489271057976617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,8192,0.1490888860490587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,10240,0.21519554985894096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,51200,0.9354186587863498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,7168,0.05000888970163134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,7168,0.13098133934868708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,10240,0.06880444288253784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,5120,0.10644621981514825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,6144,0.11609955628712972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,8192,0.15331466992696127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,5120,0.0980444418059455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,5120,0.03776444329155816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,5120,0.09949422544903225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,7168,0.13361067242092556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,4096,0.03223911258909438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,4096,0.07966755496131049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,6144,0.043736000855763756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,12288,0.24651734034220377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,3584,0.07087200217776828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,3584,0.07311199771033393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,4096,0.08596889177958171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,6144,0.10904000202814738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,3072,0.06286844280030993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,6144,0.11704799864027236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,3072,0.06515377759933472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,3584,0.029459555943806965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,2560,0.05384000142415365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,3072,0.06255288918813069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,2048,0.04036088784535726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,2560,0.023931556277804907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,2560,0.056335111459096275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,2048,0.020594666401545208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,2048,0.044919110006756253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,3072,0.026968000663651362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,1536,0.01754399968518151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,1024,0.02350666622320811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,1024,0.028187556399239436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,3584,0.07832177480061848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,4096,0.08306044340133667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,2560,0.049963557057910495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,768,0.02162488963868883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,1024,0.014865777558750577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,1536,0.03655822078386942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,768,0.013797333670987023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,1536,0.039665778477986656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,768,0.02732533382044898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,512,0.012950222525331708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,256,0.013606222139464485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,512,0.022658665974934895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,1536,0.03161688976817661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,2048,0.04822044571240743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,1024,0.031679110394583807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,256,0.017160000072585214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,256,0.01772622267405192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,512,0.0177511109246148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,768,0.02293866707219018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,128,0.012090666426552666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,32,0.011715555356608497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,512,0.019239111079110038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,128,0.016192888220151264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,32,0.016196444630622864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,128,0.012065778175989786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,256,0.011854222251309289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,64,0.011077333655622272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,128,0.016174222032229107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,65536,0.3802186648050944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,64,0.016196444630622864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,65536,1.316515604654948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,16384,0.2633048958248562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,51200,1.0075919893052843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,16384,0.2761466768052843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,65536,1.0710852940877278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,16384,0.09687911139594184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,12288,0.2068071100446913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,51200,0.740199089050293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,12288,0.07459111346138848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,10240,0.1802417834599813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,51200,0.29879734251234263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,12288,0.19060088528527153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,10240,0.06146577994028727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,10240,0.16735022597842744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,10240,0.1596062183380127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,16384,0.24969421492682564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,8192,0.1364604499604967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,7168,0.12591377894083658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,8192,0.13203733497195774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,12288,0.20858755376603869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,7168,0.04508444335725573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,7168,0.12021422386169434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,7168,0.11602489153544109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,8192,0.14540444480048284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,6144,0.10366577572292751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,6144,0.10177333487404717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,65536,1.1984017690022786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,8192,0.050492445627848305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,5120,0.034485333495669894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,4096,0.0745119982295566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,5120,0.08737422360314263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,4096,0.07117688655853271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,3584,0.06508888800938924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,5120,0.08268266916275024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,51200,0.7928444544474283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,3584,0.06358844704098172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,3072,0.0535715553495619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,3584,0.06490311357710096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,6144,0.09635200103123982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,3072,0.05627644724316067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,4096,0.02920266654756334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,2560,0.04464977648523119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,3072,0.057796445157792836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,6144,0.03920000129275852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,2560,0.022008889251285132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,2560,0.04837333493762546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,2560,0.050402667787339955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,3584,0.02676533328162299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,2048,0.01941155559486813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,2048,0.04329866502020094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,3072,0.024275556206703186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,1536,0.033576889170540705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,1536,0.036024888356526695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,1024,0.021451556020312842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,5120,0.08796711100472344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,1536,0.02737688852681054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,1024,0.014276444911956787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,768,0.017310221989949543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,4096,0.07227733400132921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,1536,0.016757332616382174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,2048,0.03523555397987366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,512,0.013271110753218332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,768,0.01313333296113544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,768,0.024157333705160353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,2048,0.040433777703179255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,512,0.017215111189418368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,512,0.020848888489935134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,1024,0.029620445436901514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,256,0.015466666883892484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,128,0.009343999955389235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,256,0.016320889194806416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,128,0.014847111370828418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,64,0.008372444245550368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,128,0.011622221933470832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,256,0.010672888822025724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,32,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,64,0.014799111419253878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,32,0.014866666661368476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,256,0.011413333316644033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,768,0.020800888538360596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,65536,0.9397395451863607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,65536,0.3526791201697455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,51200,0.677376005384657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,65536,1.1479786766899955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,16384,0.27183466487460667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,51200,0.2795902093251546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,51200,0.8564089139302572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,512,0.012116444607575735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,51200,0.6758657561408149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,16384,0.2317031224568685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,12288,0.17938400639428032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,16384,0.2157848940955268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,12288,0.0651057759920756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,12288,0.18243911531236437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,12288,0.16482222080230713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,65536,0.9166506661309136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,10240,0.14812621805402967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,10240,0.14095377922058105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,16384,0.0848346683714125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,8192,0.11954311529795329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,10240,0.1685875521765815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,7168,0.10538844267527263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,1024,0.02569688856601715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,7168,0.03996977872318692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,8192,0.045047110981411405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,6144,0.10225155618455674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,8192,0.11635643906063503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,6144,0.03562488820817735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,6144,0.08885600169499715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,5120,0.08372000190946792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,10240,0.05644977755016751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,128,0.014992000328169929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,4096,0.06359644730885823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,5120,0.07766577932569715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,5120,0.07808888620800443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,7168,0.10079733530680339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,3584,0.05758044454786512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,7168,0.11886044343312581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,4096,0.06390844451056586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,6144,0.09187733464770848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,3584,0.056495997640821666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,8192,0.1361288891898261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,5120,0.03037244412634108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,3072,0.049916442897584706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,2560,0.040780444939931236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,3072,0.05087288882997301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,4096,0.06326933039559259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,2560,0.04321600000063578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,2560,0.019721777902709115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,2560,0.0446684459845225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,3072,0.04940355486339993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,3584,0.05699466996722751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,1536,0.024679111109839544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,2048,0.01703644461101956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,2048,0.03635644581582811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,2048,0.03822044531504313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,1024,0.017599110802014668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,3072,0.021550221575631037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,1536,0.03208622336387634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,4096,0.026380444566408794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,768,0.014900444282425774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,1024,0.013183111117945777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,2048,0.02999733222855462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,1024,0.025664000047577754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,3584,0.02367377777894338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,768,0.012160888976520963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,768,0.01869066721863217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,768,0.021529777182473078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,256,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,512,0.015832000308566622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,512,0.018452444010310702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,1536,0.02979288829697503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,256,0.013909333282046847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,128,0.008107555409272512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,1024,0.022707555029127333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,512,0.011731555892361535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,128,0.013435555828942193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,64,0.00720000018676122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,128,0.01075466639465756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,128,0.014228444960382251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,32,0.007289778027269576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,512,0.011378666592968835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,1536,0.014853333433469137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,64,0.013461333182122974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,32,0.013456888496875763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,256,0.010785777535703448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,256,0.0158924443854226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,65536,0.7669448852539062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,51200,0.5535306930541992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,65536,0.9212267133924695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,16384,0.19776977433098686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,51200,0.696382204691569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,65536,0.7150790956285266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,51200,0.557021352979872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,16384,0.06901066833072238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,12288,0.13693511486053467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,16384,0.18646844228108725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,65536,0.27564533551534015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,12288,0.054357333315743335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,12288,0.1621840000152588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,12288,0.13645066155327693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,8192,0.08475377824571396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,10240,0.1283831066555447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,51200,0.2339155673980713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,10240,0.11635911464691162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,7168,0.0904044442706638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,8192,0.10267821947733562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,8192,0.09615200095706516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,16384,0.19803555806477866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,7168,0.0333555539449056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,7168,0.08627288871341282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,10240,0.10538933012220596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,6144,0.07943822277916802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,6144,0.02966488732231988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,10240,0.04656889041264852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,6144,0.07491466734144422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,4096,0.04797066582573784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,5120,0.06743288702434964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,5120,0.06502399841944377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,4096,0.022657778528001573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,4096,0.055049777030944824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,3584,0.04288355509440104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,5120,0.05383200115627713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,3584,0.049470222658581205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,3072,0.03706666827201843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,5120,0.02644444505373637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,3072,0.043974220752716064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,3072,0.01909244391653273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,2560,0.031493332650926374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,3584,0.04895822207132975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,2560,0.03790666659673055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,8192,0.03714399867587619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,2560,0.017216889394654166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,2560,0.038322667280832924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,3072,0.04383111000061035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,2048,0.01534400052494473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,2048,0.03330488999684652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,1536,0.019581332802772522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,3584,0.02100444502300686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,1536,0.026309332913822595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,1536,0.013807111316257052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,1024,0.01495911180973053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,1024,0.020026667250527274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,1024,0.02236533330546485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,768,0.012297777665985955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,1024,0.011794666449228922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,768,0.01679999960793389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,768,0.019127999742825825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,2048,0.02439022229777442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,6144,0.0656942195362515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,512,0.014138667119873894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,512,0.010421333213647207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,256,0.007548444800906711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,512,0.016574222180578444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,256,0.012648888760142855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,256,0.014371555712487964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,256,0.009908444351620143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,4096,0.055664002895355225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,128,0.012133333418104382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,128,0.013461333182122974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,64,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,1536,0.02773511078622606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,32,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,64,0.012083555261294046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,128,0.006948444578382704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,7168,0.0911253359582689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,32,0.012083555261294046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,65536,0.6349066628350152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,128,0.010034666293197209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,65536,0.25606489181518555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,768,0.011057777537239922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,51200,0.509901311662462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,65536,0.6999404695298937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,51200,0.5884933471679688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,16384,0.1667004426320394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,51200,0.5467697779337565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,16384,0.1686391168170505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,16384,0.063737776544359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,12288,0.1265031099319458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,12288,0.12627555264367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,12288,0.1338826682832506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,12288,0.048471109734641186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,10240,0.1117511060502794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,10240,0.10650133424335056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,512,0.00943911075592041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,10240,0.11355733871459961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,8192,0.08616177903281318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,10240,0.04246222310596042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,2048,0.032163557079103254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,8192,0.034394668208228216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,7168,0.0741199983490838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,8192,0.08618577983644272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,8192,0.0941315558221605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,7168,0.03170222375128005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,7168,0.07675555679533216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,7168,0.08282933632532756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,65536,0.7408711115519205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,6144,0.06646044386757745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,6144,0.07260533173878987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,51200,0.20429689354366729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,5120,0.056863109270731606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,5120,0.02461244497034285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,5120,0.06250666909747653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,4096,0.02146222194035848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,4096,0.046910223033693105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,3584,0.041119111908806696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,4096,0.05199199914932251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,16384,0.17691732777489555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,3584,0.019558222757445443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,3584,0.04228444563017952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,3584,0.04711822337574429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,3072,0.017996443642510306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,3072,0.04257955484920078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,6144,0.06574044624964397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,2560,0.03250222073660956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,5120,0.05388711227311028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,6144,0.028023110495673284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,3072,0.035051554441452026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,2048,0.027622222900390625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,2560,0.03032533327738444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,2048,0.03133599956830343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,1536,0.020061333974202473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,2048,0.014701333310869006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,1536,0.013247999880048962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,1024,0.015156444576051501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,1536,0.026515555050637987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,2560,0.016435555285877652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,2560,0.03709777858522203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,768,0.01262399968173769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,1024,0.011414222419261932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,1024,0.021682666407691106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,768,0.014501333236694336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,512,0.009363555245929295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,768,0.018499554859267343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,1536,0.022713777091768052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,512,0.012452444268597497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,256,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,512,0.015882665912310284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,2048,0.02531911101606157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,1024,0.01772799922360314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,256,0.009660444325870937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,256,0.011172444456153445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,4096,0.04641866683959961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,256,0.012979555461141797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,64,0.00739733295308219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,768,0.010767111347781287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,128,0.010748444332016839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,32,0.01015733347998725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,128,0.012109333442317115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,64,0.010663111176755695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,32,0.010460444622569613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,512,0.010023111270533668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,65536,0.5380729039510092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,128,0.00675733346078131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,65536,0.2503022352854411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,65536,0.6481946839226617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,128,0.009696000152164036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,16384,0.13494577672746447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,65536,0.5501262346903483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,51200,0.20018488830990264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,12288,0.1065431104765998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,16384,0.06501866711510552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,16384,0.1484151151445177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,51200,0.4330631097157796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,16384,0.14469422234429255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,10240,0.08839199940363567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,12288,0.11249955495198567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,3072,0.037399109866884016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,51200,0.42066489325629336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,10240,0.04287911123699612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,10240,0.09290755457348293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,51200,0.5213600264655219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,8192,0.03388355506791009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,8192,0.07830222447713216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,7168,0.060236447387271456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,8192,0.0792542232407464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,12288,0.048708445496029325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,7168,0.06951555278566149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,6144,0.0558097759882609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,12288,0.10992355479134454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,10240,0.09518311421076457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,8192,0.07186133331722684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,5120,0.043750223186280995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,6144,0.02784444557295905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,6144,0.06076266368230184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,5120,0.024145777026812237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,5120,0.05132888754208883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,5120,0.052820444107055664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,7168,0.031128889984554712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,3584,0.030184000730514526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,7168,0.06900088654624091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,6144,0.059975114133622914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,3584,0.03829244441456265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,3584,0.03975733452373081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,4096,0.036419557200537786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,3072,0.033907555871539645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,4096,0.042378667328092784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,3072,0.03584533267550998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,2560,0.02293511066171858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,3072,0.017481777403089736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,2048,0.018361777067184448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,2560,0.014745778507656522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,2560,0.029714667134814795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,2560,0.031585776143603854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,2048,0.02518400053183238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,3072,0.02589955594804552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,2048,0.02772444486618042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,1536,0.012006222373909421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,1536,0.02274488906065623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,1024,0.011205332974592844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,4096,0.04390133420626322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,1024,0.010383110907342698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,1024,0.015847999188635085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,1024,0.018589332699775696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,1536,0.014913777510325113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,2048,0.013999111122555204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,512,0.008314666648705801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,768,0.010026666853162978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,768,0.01628800067636702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,1536,0.0207306659883923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,4096,0.02085955606566535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,256,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,512,0.009407999614874521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,256,0.009022222624884712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,256,0.010423111418883005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,128,0.005975111077229182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,256,0.012773333324326409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,3584,0.018852444158660043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,128,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,64,0.005688888745175467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,32,0.005752000129885144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,128,0.011791999969217511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,64,0.010051555931568146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,512,0.011426666544543372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,32,0.010020444790522257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,512,0.014466666513019137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,65536,0.5009279780917698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,768,0.009706666900051964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,65536,0.2085262272093031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,65536,0.586102220747206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,128,0.010089777410030365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,51200,0.16507466634114584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,16384,0.1283128923839993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,51200,0.47491200764973956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,12288,0.09940889146592881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,51200,0.4275102350446913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,16384,0.1330275535583496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,16384,0.14021955596076116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,12288,0.04197688897450765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,10240,0.0813351141081916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,51200,0.3474284542931451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,12288,0.1091466678513421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,8192,0.0588773356543647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,10240,0.03621777892112732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,10240,0.08666044473648071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,65536,0.5528728697035048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,10240,0.09177688757578532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,7168,0.05763733386993408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,8192,0.0292559994591607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,8192,0.07016977998945448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,8192,0.07628444168302748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,7168,0.06244266695446438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,7168,0.06781333022647433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,12288,0.10238489177491929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,5120,0.04284088810284933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,6144,0.05987199809816149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,768,0.013459555804729462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,5120,0.04667377803060743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,5120,0.02137600051032172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,4096,0.034509334299299456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,16384,0.05273866653442383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,5120,0.051776889297697276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,6144,0.047305779324637524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,4096,0.018075555562973022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,3584,0.027447111076778833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,4096,0.04348799917432997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,6144,0.024029332730505202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,3584,0.016728889611032274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,3072,0.025536000728607178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,3584,0.0391244457827674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,3072,0.01551466683546702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,3072,0.03148444493611654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,3072,0.03534577621353997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,7168,0.026683555708991155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,4096,0.03946311275164286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,2048,0.01754844519827101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,2560,0.03072711163096958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,6144,0.05429777834150526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,3584,0.0350835555129581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,1536,0.014137778017255994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,2048,0.022895110978020564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,2048,0.027069333526823256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,2560,0.021691555778185528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,1536,0.019289778338538278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,1024,0.010382222632567087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,1536,0.022358222140206233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,2560,0.027201778358883325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,1024,0.010372444159454769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,768,0.00872622181971868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,1024,0.018072888255119324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,512,0.0069751110341813825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,768,0.009718221922715506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,2048,0.013280889226330651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,768,0.016173332929611206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,256,0.006088888893524806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,512,0.01074577785200543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,512,0.009355555805895064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,512,0.013507555756303998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,128,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,256,0.009730666875839233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,256,0.01216533366176817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,2560,0.014364444547229342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,64,0.004896000027656555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,128,0.009392889009581672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,32,0.004957333207130432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,128,0.011101333631409539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,768,0.01256711118751102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,64,0.009343110852771336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,32,0.009356444080670675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,65536,0.385700437757704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,1536,0.011735999749766456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,256,0.008815111385451423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,65536,0.1620417833328247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,65536,0.49350134531656903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,1024,0.014949333336618213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,16384,0.09870666927761501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,65536,0.40422667397393125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,51200,0.38934755325317383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,16384,0.041638222005632185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,51200,0.31490310033162433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,16384,0.10993066761228774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,12288,0.03324977888001336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,12288,0.08295200268427531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,128,0.0087013335691558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,10240,0.07579288880030315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,12288,0.07855466339323255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,16384,0.11689599355061848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,8192,0.05139644609557258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,10240,0.0706826647122701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,12288,0.0879324475924174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,8192,0.0615288880136278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,7168,0.04361777835422092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,8192,0.059798220793406166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,7168,0.021552888883484736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,51200,0.12917599413130018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,6144,0.039981332090165876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,7168,0.05303288830651177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,10240,0.02922044528855218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,6144,0.019368888603316415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,5120,0.031685332457224526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,6144,0.04786933461825053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,6144,0.04735555582576328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,4096,0.02606666584809621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,8192,0.023507555325826008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,5120,0.04091377721892463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,51200,0.2793786790635851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,5120,0.041136887338426374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,3584,0.022153778208626643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,4096,0.015211555692884656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,4096,0.034156444999906756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,3072,0.02047911120785607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,3584,0.014084445105658637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,3584,0.030744890371958416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,3584,0.03158400124973721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,2560,0.01724799970785777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,3072,0.013207111093733044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,3072,0.027261333333121404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,3072,0.028544889556037054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,2048,0.014912888407707214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,5120,0.017519111434618633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,2560,0.02388444377316369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,10240,0.06176977687411838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,2560,0.025386666258176167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,1536,0.012187555432319641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,2048,0.011007111105653973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,2048,0.02033866610791948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,2048,0.02165866725974613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,1024,0.009574222067991892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,1536,0.01686044368478987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,1536,0.018049778209792245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,1024,0.013099555340078143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,768,0.008345777789751688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,1024,0.01550222271018558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,768,0.01142133358452055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,768,0.013775111072593264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,768,0.008792888787057664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,512,0.009704888694816166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,1024,0.009391999906963771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,2560,0.012097777591811286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,512,0.008357333640257517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,256,0.005590222362014983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,512,0.012122666670216454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,256,0.009045333498054082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,512,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,128,0.008664888640244802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,128,0.008031110796663497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,64,0.004271999829345279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,4096,0.03472266594568888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,32,0.004606222112973531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,64,0.00868266655339135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,32,0.008683555656009251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,65536,0.2912355528937446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,7168,0.05489155650138855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,1536,0.010117332968446944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,65536,0.3936426639556885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,51200,0.23814667595757377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,128,0.010694222317801582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,51200,0.3383839925130208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,51200,0.11979199780358209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,16384,0.0808159973886278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,51200,0.3100702232784695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,256,0.01073333372672399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,16384,0.10268977615568374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,12288,0.07060177458657159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,16384,0.10530310869216919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,12288,0.07898666461308797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,12288,0.08181155390209623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,256,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,128,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,16384,0.04032177726427714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,8192,0.04543377624617683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,65536,0.15387645032670763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,10240,0.06979822450213961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,12288,0.03194133440653483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,8192,0.022944000032213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,8192,0.05465244584613376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,8192,0.058543112542894155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,7168,0.02103022237618764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,7168,0.04832888974083794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,6144,0.034599112139807806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,10240,0.028872887293497723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,6144,0.042224887344572276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,5120,0.02900800108909607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,6144,0.01926488843229082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,10240,0.05535822113355001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,5120,0.03653511073854234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,5120,0.016752888758977253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,5120,0.040052443742752075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,65536,0.434328900443183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,7168,0.05227022369702657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,3584,0.022290666898091633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,4096,0.014858666393491956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,4096,0.03104355600145128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,4096,0.03396888905101352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,3072,0.019727999965349834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,3584,0.013795555465751223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,6144,0.04617599977387322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,3584,0.03061244553989834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,3072,0.012770666844315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,3072,0.0251893334918552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,3072,0.028014222780863445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,4096,0.024367999699380662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,10240,0.06682666805055407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,2560,0.02159022291501363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,2560,0.011764444410800934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,2560,0.02429955535464817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,1536,0.0116462219092581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,2048,0.010758221977286868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,2048,0.01827111177974277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,2048,0.02136444383197361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,3584,0.027887112564510767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,1536,0.015439111325475903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,7168,0.039692444933785334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,1536,0.01807466646035512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,768,0.00833155545923445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,1024,0.009351111120647853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,1024,0.012476444244384766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,2048,0.014911999305089315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,512,0.007024889190991719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,768,0.010404444403118556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,768,0.013791110780504016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,512,0.008359111017651027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,1024,0.009350222018029954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,256,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,2560,0.017104888955752056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,256,0.008359111017651027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,1024,0.014914666612943014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,128,0.004947555561860402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,256,0.010393777655230628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,256,0.00798044436507755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,128,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,64,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,512,0.0090524446633127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,128,0.01038933296998342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,32,0.0053048887186580235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,64,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,65536,0.262172434065077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,32,0.007710222568776872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,768,0.00904799997806549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,65536,0.14540799458821616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,51200,0.18543467256757948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,128,0.008009778128729926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,65536,0.3976302146911621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,1536,0.010059555371602377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,16384,0.06722044282489352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,51200,0.26647644572787815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,51200,0.11549066172705756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,12288,0.04890844557020399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,16384,0.037754668129815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,16384,0.08595822254816692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,16384,0.10432889064153035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,10240,0.03999555442068312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,12288,0.0643848909272088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,65536,0.3374364376068115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,8192,0.030913776821560327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,512,0.011874666644467248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,10240,0.06895555390252008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,8192,0.045031110445658364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,7168,0.02773155603143904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,51200,0.3109955522749159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,8192,0.058133330610063344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,12288,0.030788444810443457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,7168,0.04046577877468533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,7168,0.05111644334263272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,10240,0.054851555162005954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,6144,0.03548177745607164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,6144,0.017948443690935772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,5120,0.02047288914521535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,6144,0.04545066754023234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,8192,0.02204622162712945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,4096,0.017644445101420086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,5120,0.016255110502243042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,5120,0.03911822372012668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,7168,0.02032355633046892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,4096,0.02548355526394314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,4096,0.014482667048772177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,4096,0.033032000064849854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,10240,0.027775999572541978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,3072,0.01406133340464698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,3584,0.01350933313369751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,3584,0.02951822347111172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,3072,0.020364445116784837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,3072,0.012482666307025485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,3072,0.02681333323319753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,6144,0.025646222962273493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,12288,0.08028799957699247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,2560,0.023575110567940608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,2048,0.010764444039927589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,3584,0.023040889037979975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,2048,0.015463999576038785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,2048,0.010659555594126383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,2048,0.020655999581019085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,2560,0.012252444194422828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,5120,0.03050666716363695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,1024,0.007634667058785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,1536,0.013100444442696042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,1536,0.01755288905567593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,2560,0.018009778526094224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,768,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,1024,0.010433778166770935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,2560,0.011691555380821228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,1024,0.015110222829712762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,512,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,768,0.00868977771864997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,768,0.009023110899660323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,1536,0.008990222381220924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,256,0.004661333229806689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,512,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,512,0.0118568887313207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,1536,0.009855999714798397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,256,0.007371555599901412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,3584,0.016359999775886536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,256,0.01072533345884747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,1024,0.008922666311264038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,64,0.0038951109680864546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,128,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,32,0.0038657776183552216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,128,0.010383110907342698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,768,0.013449778159459433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,64,0.0069919998447100324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,512,0.008286221987671321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,32,0.007299555672539606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,65536,0.10009421904881795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,128,0.0042071110672420925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,256,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,65536,0.2859111097123888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,65536,0.24507912000020346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,16384,0.05017333229382833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,128,0.00701155596309238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,51200,0.22073777516682944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,51200,0.193451550271776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,65536,0.18512888749440512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,16384,0.07230044735802545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,16384,0.06902399990293714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,12288,0.022258667482270136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,12288,0.05329777797063192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,51200,0.07954044474495782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,10240,0.046572443511750966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,8192,0.024324445260895625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,12288,0.0380639996793535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,16384,0.0274008893304401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,10240,0.0458586679564582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,7168,0.021703999903466966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,8192,0.016134222348531086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,8192,0.03849600089920892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,12288,0.05556977457470364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,8192,0.038853334056006536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,6144,0.019768888751665752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,7168,0.014704000618722705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,7168,0.034252444903055825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,5120,0.016833777228991192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,6144,0.013591999808947245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,6144,0.030508445368872747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,6144,0.03141866789923774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,10240,0.020277332928445604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,5120,0.02590222160021464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,5120,0.026886221435334947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,4096,0.01107911103301578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,3584,0.012753778033786349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,51200,0.12615466117858887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,7168,0.035049776236216225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,3584,0.019852444529533386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,3072,0.011681777735551199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,3584,0.010774222513039907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,4096,0.01418133411142561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,5120,0.012416000167528788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,2560,0.010382222632567087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,3072,0.010381333529949188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,3072,0.017860444055663217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,3072,0.019362666540675692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,2048,0.009053332938088311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,4096,0.02311733365058899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,2560,0.01585422290696038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,10240,0.029450668228997126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,2560,0.017235555582576327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,2048,0.009413333402739631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,2048,0.013790222505728403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,2048,0.015429332852363586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,1024,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,1536,0.008709333009190029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,1536,0.013612444202105204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,1024,0.008347555167145198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,4096,0.02165422174665663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,2560,0.009676444033781687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,1536,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,768,0.008319999608728621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,512,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,768,0.010737777584128909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,3584,0.021136000752449036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,512,0.007603555917739868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,256,0.0042364444169733255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,1024,0.009460444251696268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,1024,0.012087999946541257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,256,0.006671999891599019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,256,0.007333333293596904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,128,0.00388355553150177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,256,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,768,0.007747555772463481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,64,0.0037644443412621817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,128,0.008968888885445064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,32,0.003887999802827835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,512,0.007297777467303806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,512,0.009701333112186855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,64,0.006575111299753189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,65536,0.14621955818600124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,768,0.005649777750174205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,32,0.0063235556913746735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,51200,0.1183004379272461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,65536,0.09827378061082627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,65536,0.24031466907925078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,65536,0.24221955405341256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,51200,0.07913688818613689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,51200,0.18923377990722656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,128,0.007295110987292395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,51200,0.19219288561079237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,16384,0.02693866689999898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,16384,0.062486224704318576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,16384,0.06748000118467543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,10240,0.023736889163653057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,12288,0.04798044429885017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,12288,0.0527671111954583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,10240,0.020061333974202473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,10240,0.04087200098567539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,16384,0.038789331912994385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,8192,0.01941155559486813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,128,0.0063075555695427795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,12288,0.027951111396153767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,7168,0.018147556318177115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,8192,0.033944000800450645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,12288,0.021889777647124395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,8192,0.03762844536039565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,6144,0.015999111864301894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,7168,0.014885332849290637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,10240,0.0454711119333903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,6144,0.02681688964366913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,5120,0.013943110903104147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,8192,0.016199999385409884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,5120,0.022857778602176245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,7168,0.03035466538535224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,5120,0.02658844490845998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,4096,0.019704000817404855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,6144,0.0136817776494556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,4096,0.011123555401961008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,4096,0.023051554958025616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,4096,0.011780444118711682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,3584,0.017755554782019723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,3584,0.020853334003024634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,3584,0.010713777608341642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,1536,0.011767999993430244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,6144,0.03038933210902744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,3072,0.016120000018013846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,3072,0.01891911029815674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,3584,0.011045333411958484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,2560,0.014130666851997375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,2560,0.009859555297427708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,5120,0.01237066669596566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,3072,0.010013333625263637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,2048,0.012358222570684222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,1536,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,3072,0.010054222411579555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,1536,0.01068800025516086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,1536,0.008681777450773451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,2048,0.007628444168302748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,7168,0.03388444582621256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,1024,0.008692444198661381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,2048,0.009229333036475712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,1024,0.011505777637163797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,768,0.005471111171775394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,1024,0.008088889221350351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,768,0.007947555846638149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,768,0.007640000018808577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,512,0.00490044429898262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,768,0.010760889285140567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,2560,0.01715288890732659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,256,0.003896000070704354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,512,0.009768889182143742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,2048,0.0151582227812873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,256,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,128,0.0038737778862317405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,2560,0.008923555413881937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,256,0.006980444822046492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,128,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,128,0.005965333431959152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,64,0.0032515554792351196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,128,0.009018667042255402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,1024,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,512,0.007480888730949826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,65536,0.10577066739400227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,64,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,32,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,51200,0.08169333140055339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,65536,0.20046044720543754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,256,0.00867377801073922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,65536,0.2323333422342936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,16384,0.0328746669822269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,51200,0.0790897806485494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,1536,0.01311644415060679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,32,0.003583111282851961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,51200,0.18351111147138807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,512,0.006754666566848755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,16384,0.052988443109724254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,16384,0.06480622291564941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,12288,0.021314667330847845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,12288,0.04122755593723721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,65536,0.09789422485563491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,12288,0.05022933416896396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,8192,0.017101332545280457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,10240,0.03513066636191474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,10240,0.04335911075274149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,8192,0.02929688824547662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,8192,0.015522667103343539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,16384,0.026543999711672466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,10240,0.020641777250501845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,7168,0.02647200061215295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,6144,0.013999111122555204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,7168,0.03262844350602891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,7168,0.014151111245155334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,10240,0.019319110446506076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,5120,0.01201600001917945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,6144,0.013112000293201871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,51200,0.15955289204915366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,6144,0.028816001282797918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,12288,0.024142222272025213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,4096,0.010292444792058732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,5120,0.024918221765094336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,4096,0.017214222086800467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,4096,0.02165066699186961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,3584,0.009462222456932068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,4096,0.01072622256146537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,6144,0.023196443915367126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,3584,0.015507555670208402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,5120,0.01199288914601008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,3584,0.019713777634832595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,8192,0.036087999741236367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,3072,0.00979466653532452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,3072,0.014134221606784396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,3072,0.018019555343521964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,2560,0.012517333030700684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,2560,0.016233777006467182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,2560,0.009648000200589498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,2048,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,7168,0.01538222200340695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,2048,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,2560,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,1536,0.009714666340086196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,1536,0.008361777497662438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,1024,0.005608888963858287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,5120,0.02016888890001509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,1024,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,2048,0.014504889647165934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,1024,0.007976888782448238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,1536,0.012708444562223224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,3072,0.008664000365469191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,768,0.01038044442733129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,768,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,1024,0.011132444772455426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,512,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,512,0.007296000089910295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,512,0.00925066653225157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,256,0.003937777959638172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,3584,0.010422222316265106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,256,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,256,0.007015110717879401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,256,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,768,0.004952888935804367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,128,0.005583111196756363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,64,0.003572444534964032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,768,0.0070471109615431885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,32,0.0035502223504914176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,64,0.00565244464410676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,65536,0.060356444782680936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,32,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,512,0.004263111286693149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,65536,0.09758044613732232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,51200,0.04982044299443563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,65536,0.2289173338148329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,51200,0.1262800031238132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,1536,0.00700800038046307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,128,0.0038817777401871155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,65536,0.15663022465176052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,128,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,16384,0.02517066730393304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,12288,0.01832266648610433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,2048,0.01108000013563368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,51200,0.07851999998092651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,16384,0.019188443819681805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,10240,0.016172443826993305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,12288,0.02091022166940901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,128,0.00833155545923445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,10240,0.028143111202451918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,10240,0.019273777802785236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,10240,0.04229866796069675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,16384,0.063664886686537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,8192,0.015438222222858004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,7168,0.011936888926559024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,16384,0.04194044404559665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,8192,0.01314933349688848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,7168,0.03153422143724229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,7168,0.014308444327778287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,8192,0.023629332582155865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,8192,0.03548977772394816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,6144,0.018941332896550495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,5120,0.009420444567998251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,6144,0.028185778194003638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,5120,0.016752888758977253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,5120,0.02459822263982561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,4096,0.008240888516108194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,51200,0.1808640029695299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,12288,0.03266400098800659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,4096,0.014491554763582019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,6144,0.013125333521101209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,12288,0.049343112442228526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,3584,0.007358222372002072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,3584,0.013200889031092325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,3584,0.010263111028406356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,3584,0.019527110788557265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,5120,0.011791110866599612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,3072,0.012098666694429187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,2560,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,3072,0.018057778477668762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,7168,0.02122755514250861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,4096,0.02104533380932278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,2560,0.010826666322019366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,2560,0.009391999906963771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,2560,0.016082665986484952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,6144,0.01070577816830741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,2048,0.00964355551534229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,2048,0.01463822192615933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,4096,0.010737777584128909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,1536,0.008492444124486711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,1536,0.008403555386596257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,3072,0.00701333334048589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,1024,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,1536,0.004937777916590373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,1024,0.008045333127180735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,768,0.003936888857020272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,1024,0.011237333218256632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,2048,0.005591111050711737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,1536,0.012495111260149213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,512,0.0035208890007601846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,768,0.010051555931568146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,2048,0.008948444492287105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,512,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,512,0.009353777600659264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,256,0.0053004444473319584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,256,0.008356444537639618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,256,0.0069226668112807805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,768,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,128,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,3072,0.009725333087974125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,256,0.0032035555276605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,64,0.0029128889242808023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,128,0.006959999601046245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,32,0.0031848889258172777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,65536,0.046055999067094594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,64,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,32,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,1024,0.0042462220622433555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,51200,0.03747999999258254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,65536,0.22836176554361978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,512,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,128,0.008040000167157915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,16384,0.012474666866991254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,51200,0.07833511299557157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,51200,0.18049689133961996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,16384,0.041823112302356295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,768,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,16384,0.06296177705128987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,65536,0.15659199820624456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,12288,0.020626667473051283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,12288,0.032495998673968844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,10240,0.011587555209795633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,12288,0.04894222153557671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,51200,0.12313777870602077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,8192,0.011114666859308878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,10240,0.018968888454967074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,10240,0.027671110298898485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,10240,0.041454222467210554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,7168,0.011630222201347351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,128,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,8192,0.03465866711404588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,65536,0.09753333197699653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,6144,0.010261333651012844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,7168,0.014138667119873894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,7168,0.031376888354619346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,6144,0.018607111440764535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,6144,0.013119999733236102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,5120,0.009121778110663096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,6144,0.027851555082533095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,5120,0.011740444435013665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,8192,0.015252444479200574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,5120,0.016511999898486666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,5120,0.024078221784697637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,4096,0.010747555229398938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,7168,0.020974222156736586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,3584,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,4096,0.021146666672494676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,3584,0.010213333699438307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,3584,0.01312888910373052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,3072,0.006613333192136552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,3584,0.018944000204404194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,8192,0.023176888624827068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,4096,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,2560,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,3072,0.01760444376203749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,2560,0.009354666703277165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,2560,0.010828444527255164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,2560,0.01594311164485084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,2048,0.00941777808798684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,2048,0.008763555851247575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,3072,0.009727111293209923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,1536,0.004899555610285865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,1536,0.00833777752187517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,2048,0.00536088893810908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,1536,0.012133333418104382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,1024,0.004111111164093018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,1536,0.008319999608728621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,1024,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,1024,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,1024,0.011062222222487131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,2048,0.014363555444611443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,768,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,768,0.007663999994595845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,768,0.010086221827401055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,4096,0.014429334137174817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,512,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,512,0.00906666699382994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,768,0.0035502223504914176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,512,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,16384,0.024101333485709295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,256,0.008055110772450766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,512,0.0035342222286595237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,128,0.005290666802061929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,256,0.003206222214632564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,128,0.006990222467316522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,12288,0.010728889041476779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,64,0.002557333351837264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,32,0.0026551110463009942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,256,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,128,0.0029075555503368378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,65536,0.04452177882194519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,64,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,51200,0.03521155648761325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,32,0.0052986666560173035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,16384,0.014155555102560254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,12288,0.011783999701340994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,10240,0.011056000159846412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,51200,0.12285955746968587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,12288,0.03164533442921109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,3072,0.012080888781282636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,7168,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,256,0.0069884442620807225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,6144,0.009123555488056606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,7168,0.020976000361972388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,5120,0.008357333640257517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,65536,0.15935644838545057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,4096,0.007877333296669854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,16384,0.04114222195413377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,3584,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,128,0.008049777812427944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,3072,0.006305777778228124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,8192,0.022823999325434368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,2560,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,3072,0.012126222252845764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,2048,0.0052968888647026485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,2560,0.010741333166758219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,6144,0.01861422260602315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,1536,0.004607111215591431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,1024,0.004232888834344016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,2048,0.009570666485362584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,768,0.003589333345492681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,10240,0.027360000544124182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,512,0.0035573335157500375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,1024,0.007275555696752336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,256,0.003208000005947219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,768,0.006680888848172293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,128,0.002883555574549569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,512,0.005920889063013925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,64,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,256,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,32,0.0028515555378463534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,128,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,65536,0.040545778142081365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,5120,0.016582222448454965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,51200,0.03307466705640157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,1536,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,4096,0.014183110660976835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,3584,0.01314933349688848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,12288,0.00870577742656072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,51200,0.12238666746351455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,10240,0.0081395556529363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,12288,0.031981332434548274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,8192,0.0073520003093613525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,10240,0.027472888429959614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,8192,0.010042666561073726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,32,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,6144,0.0101742222905159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,16384,0.009715555442704095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,5120,0.009688888986905416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,7168,0.020976889464590285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,4096,0.008671110702885522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,6144,0.01862044466866387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,16384,0.0414737794134352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,5120,0.01657511128319634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,3072,0.006263111200597551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,4096,0.01424888935354021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,2560,0.00591822216908137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,3584,0.013109332985348172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,2048,0.005921777751710679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,3072,0.01198755535814497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,1536,0.004579555657174853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,2560,0.010840888652536603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,8192,0.023061333431137934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,1024,0.00397777764333619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,2048,0.009406222237481011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,1536,0.008344000412358178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,512,0.0035235554807715942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,1024,0.007337777978844113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,256,0.0032319999817344877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,512,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,64,0.005045333256324132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,256,0.005283555636803309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,128,0.005007110950019625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,64,0.00253866674999396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,3584,0.008170666793982187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,768,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,32,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,64,0.004940444396601783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,128,0.0028488888508743713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,7168,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,65536,0.15596977869669595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,32,0.00489688871635331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,768,0.003543111185232798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,16384,2.0981803470187717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,16384,0.7024950981140137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,16384,2.250330607096354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,12288,0.5108435418870714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,12288,1.584646224975586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,51200,2.617044448852539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,12288,1.6368186738755968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,10240,0.42975465456644696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,10240,1.7950026194254558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,8192,0.36583378579881454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,8192,1.1500755945841472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,8192,1.0744132995605469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,7168,0.3202915456559923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,6144,0.875948429107666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,7168,0.9589306513468424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,6144,0.28130488925509983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,10240,1.2984204822116427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,6144,0.7663155661688911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,5120,0.23757600784301758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,5120,0.6860133277045356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,7168,1.2707715564303927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,5120,0.6294844415452745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,4096,0.19821155071258545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,4096,0.5052195654975044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,3584,0.47305599848429364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,3584,0.4283946620093451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,3072,0.43074488639831543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,51200,6.570079803466797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,3072,0.15266933706071642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,4096,0.5147271156311035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,2560,0.3789111243353949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,3072,0.37713156806098086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,51200,8.318086412217882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,2048,0.31873422198825413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,2048,0.11382933457692464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,1536,0.23587910334269205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,2560,0.13440088431040445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,1024,0.15806755754682753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,1536,0.09102400143941243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,768,0.12101956208546956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,1024,0.07080977492862277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,1024,0.1542088852988349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,512,0.08027288648817274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,768,0.06509333186679415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,768,0.1260062191221449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,256,0.059614221254984535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,512,0.058607108063168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,128,0.0469448897573683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,512,0.10815466774834527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,2560,0.33024711079067653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,256,0.09678755866156684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,64,0.04358844293488396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,2048,0.2592737674713135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,128,0.09054666757583618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,64,0.0905875563621521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,1536,0.20746400621202257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,32,0.09140977594587539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,256,0.05386755532688565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,3584,0.1796720027923584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,32,0.046577778127458364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,128,0.051494220892588295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,16384,1.5235457950168187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,16384,0.5391155348883735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,65536,2.436925252278646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,51200,1.950980504353841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,16384,1.756500456068251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,12288,0.4931466844346788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,12288,1.2050889333089192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,51200,4.7276357014973955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,10240,1.023498641120063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,12288,1.3487502204047308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,10240,0.3476577864752875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,65536,6.232010735405816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,8192,0.8814675543043348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,8192,0.28521598709954155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,10240,1.064835548400879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,8192,0.8170995712280273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,7168,0.7792879740397135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,6144,0.7058053546481662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,6144,0.21777866946326363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,7168,0.248252444797092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,5120,0.48650487263997394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,6144,0.601279099782308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,7168,0.7005057864718967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,5120,0.1852817800309923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,4096,0.38629688156975645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,4096,0.15901245011223689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,3584,0.4005306561787923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,3584,0.1425973309410943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,3072,0.32091466585795086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,3584,0.34387911690606016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,51200,6.499983893500434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,65536,8.71057383219401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,2560,0.23928711149427626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,2048,0.2001244492001004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,5120,0.525984870062934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,4096,0.4072435432010227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,2048,0.09200444486406113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,2048,0.2115208837721083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,1536,0.14807111687130398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,1024,0.10475733545091416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,1536,0.0707226660516527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,3072,0.3381226592593723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,1536,0.16205867131551108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,768,0.1002204418182373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,1024,0.05668266614278158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,2560,0.10860711336135864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,512,0.08019200298521253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,2560,0.2482515441046821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,768,0.09840799702538384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,512,0.04743644595146179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,256,0.07676088809967041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,3072,0.12318578031327988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,128,0.04201333390341865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,1024,0.12182754940456814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,768,0.0517119997077518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,256,0.04882311158710056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,32,0.03488355543878343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,512,0.0854568878809611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,64,0.07217688692940606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,256,0.04389955600102743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,128,0.038776000340779625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,128,0.07210489114125569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,64,0.03321244319279989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,65536,0.7966284222073026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,16384,0.44605424669053817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,51200,1.5378470950656469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,51200,0.6314586533440484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,16384,0.18102667066786024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,16384,0.5694817966885036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,12288,0.3540364371405707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,12288,0.1391813357671102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,51200,2.1616719563802085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,12288,0.4270453453063965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,10240,0.3547786606682672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,10240,0.11859199735853408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,65536,2.6201644473605685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,10240,0.32835110028584796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,8192,0.10214044650395711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,65536,2.2203405168321395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,8192,0.26102667384677464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,6144,0.21824799643622506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,7168,0.23872176806131998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,5120,0.18428177303738066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,6144,0.19764178329043916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,5120,0.0676693320274353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,5120,0.16430400477515325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,32,0.07278311252593994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,4096,0.16071644094255236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,8192,0.2638915644751655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,7168,0.2222391234503852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,4096,0.13500000370873347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,7168,0.0860826638009813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,3584,0.049069331751929395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,3584,0.11628266175587972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,3072,0.04296444522009956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,2560,0.09729244311650594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,2048,0.0748071140713162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,2560,0.03814933366245694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,2560,0.08687644534640843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,1536,0.04989511105749342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,2048,0.07220711310704549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,1536,0.026776888304286536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,1024,0.036335110664367676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,3072,0.10310310787624782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,6144,0.07772622505823772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,3072,0.10101511081059773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,768,0.031856000423431396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,1024,0.02233155568440755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,1024,0.04337244563632541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,768,0.020376000139448378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,768,0.03494577937655979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,512,0.018271999226676095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,512,0.02933244572745429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,256,0.01681422193845113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,1536,0.05726044707828098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,128,0.025075554847717285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,128,0.016884444488419425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,256,0.01738133364253574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,256,0.026200888885392085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,32,0.012084444363911947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,512,0.022794667217466567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,32,0.025037333369255066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,2048,0.03269066744380527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,128,0.014200000299347771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,65536,0.5783333248562282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,65536,1.6145644717746312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,3584,0.13531022601657444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,51200,1.2248871061537001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,64,0.02498400045765771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,16384,0.38199732038709855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,65536,2.1155618031819663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,51200,0.46056535508897567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,16384,0.14094755384657118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,12288,0.34257690111796063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,16384,0.4037626584370931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,4096,0.0543137788772583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,10240,0.2496204376220703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,51200,1.5216222339206273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,10240,0.09289244148466323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,8192,0.19535111056433785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,10240,0.24477156003316244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,8192,0.0750346647368537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,7168,0.17829244666629365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,8192,0.19860977596706816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,6144,0.14246755176120335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,7168,0.06702311171425714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,7168,0.17275910907321504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,5120,0.14063378175099692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,6144,0.06025955412122938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,6144,0.15411822001139322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,4096,0.11212089326646592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,12288,0.3297368950313992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,3584,0.09140000078413223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,4096,0.04273955689536201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,4096,0.10219555430942112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,3072,0.08184178007973565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,3584,0.08984622028138901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,2560,0.06825600067774455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,3072,0.03441333439615037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,64,0.012470222181744046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,2048,0.054791112740834556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,5120,0.05043199989530775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,5120,0.125873777601454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,2560,0.06754577822155423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,1536,0.04203022188610501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,1536,0.021916444102923077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,3584,0.038322667280832924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,1024,0.027873777680926855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,1024,0.01848088867134518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,768,0.021870222356584337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,1024,0.03460089034504361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,12288,0.10752799775865342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,512,0.018227555685573153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,2560,0.030813333061006334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,768,0.028033776415718928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,2048,0.026280888252788123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,256,0.013830222189426422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,512,0.02342844506104787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,2048,0.056107554170820445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,128,0.011573333707120685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,1536,0.04495111107826233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,256,0.0207360006040997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,128,0.014315555493036905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,32,0.010044444766309526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,128,0.020039111375808716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,3072,0.07874577575259738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,64,0.019941333267423842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,768,0.01703644461101956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,32,0.01997511088848114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,512,0.015456888410780164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,256,0.014314666390419006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,64,0.009718221922715506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,65536,0.4872746467590332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,16384,0.3387617799970839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,51200,1.0219137403700087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,16384,0.12186043792300755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,12288,0.25439733929104275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,16384,0.34016089969211155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,12288,0.09610844320721096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,65536,1.6762906180487738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,12288,0.27320888307359487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,10240,0.07849866814083524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,10240,0.20909066994984946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,8192,0.06379466586642794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,7168,0.12154666582743327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,51200,0.38590844472249347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,6144,0.12635821766323513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,7168,0.05778133206897312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,7168,0.14731021722157797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,51200,1.2019181781344943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,6144,0.05055110984378391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,10240,0.2103502220577664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,6144,0.12731466028425428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,4096,0.08815111054314508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,8192,0.15721155537499323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,3584,0.07725510994593303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,5120,0.10731910996966892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,4096,0.08724088801278009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,65536,1.3402186499701605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,3584,0.03304266598489549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,3072,0.06583022409015231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,3584,0.07708266708585951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,2560,0.05256888932651944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,3072,0.029603554142846003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,2048,0.04609688785341051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,2560,0.026542221506436665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,5120,0.11145599683125813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,2560,0.05800266398323906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,1536,0.033293333318498396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,2048,0.0229120006163915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,2048,0.04911466770701938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,1024,0.02533866630660163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,1536,0.019792889555295307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,1536,0.039064000050226845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,768,0.01891288823551602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,1024,0.016472889317406546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,1024,0.0302471121152242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,512,0.015171556009186639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,768,0.015150222513410779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,768,0.02440711193614536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,256,0.011490667031870948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,512,0.013737777868906656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,512,0.020664888951513503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,128,0.009823111196359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,256,0.01295555548535453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,256,0.018174222773975797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,4096,0.03552800085809495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,64,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,128,0.013099555340078143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,32,0.00867022242810991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,8192,0.16992178228166369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,64,0.017489777670966256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,3072,0.06731200218200684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,5120,0.042908443344963916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,65536,0.9510320027669271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,65536,0.42006577385796445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,128,0.017848889032999676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,16384,0.27611109945509166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,51200,0.8573590914408366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,65536,1.3592373530069988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,32,0.01756444407833947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,16384,0.09940799739625718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,12288,0.21556622452206084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,16384,0.27356799443562824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,10240,0.17255289024776885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,12288,0.08198222186830309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,8192,0.11651644441816543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,12288,0.2841377788119846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,7168,0.10296889146169026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,8192,0.13693866464826795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,7168,0.04758222235573662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,6144,0.10875911182827419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,7168,0.12605778376261392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,5120,0.08611022101508246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,6144,0.04169955518510607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,51200,0.33263911141289604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,6144,0.10542133119371201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,4096,0.07493333021799724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,5120,0.03633866707483927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,3584,0.06971644692950778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,4096,0.03068888849682278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,51200,0.9602915445963541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,10240,0.16968977451324463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,3072,0.05855911307864719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,3584,0.027860442797342937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,2560,0.04866577850447761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,3072,0.025124443901909724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,3072,0.05639377567503187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,2048,0.03712977634535895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,2560,0.02286044425434536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,2560,0.04799733228153653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,1536,0.02757155564096239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,2048,0.04089777668317159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,5120,0.08936711152394612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,1024,0.02123822271823883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,8192,0.05568710962931315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,1536,0.0330079992612203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,768,0.016697777642144095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,1024,0.014461333552996317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,1024,0.025109334124459162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,512,0.013100444442696042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,768,0.013226666384273104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,768,0.02082577844460805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,256,0.010048888623714447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,512,0.012480888929631976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,2048,0.019927110936906602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,128,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,10240,0.07902666595247057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,256,0.015522667103343539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,64,0.007270221908887227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,128,0.011552888486120435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,32,0.007324444750944774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,3584,0.06354133288065593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,64,0.015169777803950839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,32,0.015092444088723926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,512,0.017502221796247695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,256,0.011570666399266986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,4096,0.07186844613817003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,128,0.015156444576051501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,51200,0.7595226499769422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,16384,0.21856978204515246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,51200,0.2871031231350369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,51200,0.8383599917093912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,12288,0.18172000514136422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,16384,0.3077564504411486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,12288,0.06879822413126628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,65536,0.855439133114285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,10240,0.1586266623602973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,12288,0.19314044051700166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,10240,0.059522668520609535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,65536,1.1070355309380424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,8192,0.04766844378577339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,16384,0.09093333615197076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,7168,0.12676000595092773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,1536,0.01720177796151903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,7168,0.0425235562854343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,7168,0.10809599690967137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,6144,0.10322844319873387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,5120,0.08669955862892999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,8192,0.1515680021709866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,10240,0.1526497734917535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,5120,0.03285422258906894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,5120,0.07956000169118245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,8192,0.12219644917382134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,3584,0.061408003171284996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,4096,0.028616888655556574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,4096,0.06396355231602986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,65536,0.37238311767578125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,3584,0.02608088817861345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,3584,0.0574222207069397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,2560,0.03930222325854831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,3072,0.04987555411126879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,6144,0.0936266647444831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,2048,0.03256444467438592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,2560,0.021361778179804485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,2560,0.043319112724728055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,1536,0.02366222275627984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,2048,0.018852444158660043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,1024,0.017688888642523024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,2048,0.036377777655919395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,1536,0.02975377771589491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,768,0.014660444524553088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,1024,0.022966222630606756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,3072,0.023754666248957317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,512,0.011741333537631564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,768,0.01902666687965393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,4096,0.06131111250983345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,512,0.011875555747085147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,512,0.016116445263226826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,1536,0.0162435554795795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,256,0.011081777513027191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,128,0.007954667011896769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,3072,0.0543813341193729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,1024,0.013751111096805997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,128,0.011359111302428775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,6144,0.03777066535419888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,32,0.008773333496517604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,128,0.013798221945762634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,64,0.013807999591032663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,256,0.009343110852771336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,65536,0.3356160057915582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,256,0.014307555225160388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,65536,0.8607270982530383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,51200,0.5197457737392849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,16384,0.2330995665656196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,768,0.01275022245115704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,12288,0.15121866597069636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,16384,0.1024986637963189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,16384,0.22152800030178496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,51200,0.7887520260281033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,10240,0.12505688932206896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,12288,0.15680710474650064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,10240,0.05317066775427925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,8192,0.11014311181174384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,65536,0.9330924352010092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,51200,0.26656799846225315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,8192,0.10775822401046753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,32,0.013808888693650564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,7168,0.03767022159364488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,7168,0.0950017770131429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,12288,0.06765777534908719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,5120,0.0696915586789449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,6144,0.033542222446865506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,64,0.007080888582600488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,10240,0.14060978094736734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,4096,0.05808089176813761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,5120,0.06961866882112291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,8192,0.0425235562854343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,3584,0.049364444282319814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,4096,0.05691555473539564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,3072,0.03968177901373969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,3584,0.05090666810671488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,3072,0.020788444413079154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,6144,0.08262399832407634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,2560,0.03385688861211141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,5120,0.029456890291637842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,2048,0.027525333894623652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,2560,0.01925511161486308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,4096,0.02493688960870107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,2560,0.038875556654400296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,1536,0.022328888376553852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,2048,0.016591999265882704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,1024,0.01702311138312022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,6144,0.08391111426883274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,1536,0.026879999372694228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,768,0.014515555567211576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,1024,0.012656889028019376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,1024,0.020605333977275424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,3072,0.04487555556827121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,512,0.011945777469211154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,768,0.017252445220947266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,512,0.011076444553004371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,512,0.014535999960369535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,256,0.008837333156002892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,3584,0.02281066609753503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,256,0.010433778166770935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,256,0.012799111505349478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,128,0.01037599974208408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,128,0.012534221841229333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,64,0.0069520001610120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,768,0.011745777395036487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,32,0.006980444822046492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,64,0.012418666647540199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,32,0.012439111040698158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,2048,0.033129778173234724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,128,0.007345777418878343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,65536,0.6844835811191134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,65536,0.2694773409101698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,51200,0.43621333440144855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,16384,0.16432266765170628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,51200,0.24289245075649687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,65536,0.7960435549418131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,12288,0.13421689139472112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,16384,0.07226933373345269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,51200,0.6191724671257867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,16384,0.18745688597361246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,12288,0.06285066737069024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,1536,0.014456000592973499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,8192,0.08663911289638943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,10240,0.04519822200139364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,7168,0.07376266850365533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,10240,0.11288889249165852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,8192,0.03557422094874912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,8192,0.09281511439217462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,7168,0.032229334115982056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,5120,0.05438489052984449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,6144,0.028805332051383123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,6144,0.07030933433108859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,4096,0.0472871098253462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,7168,0.09425066577063666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,5120,0.060303999318016895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,10240,0.1095599995719062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,3584,0.03818666603830125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,4096,0.02223199937078688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,12288,0.13560356034172907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,4096,0.04912177721659342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,3584,0.02027555637889438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,2560,0.026925333672099646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,3072,0.018428444862365723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,3072,0.03899644480811225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,7168,0.08182044161690606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,2048,0.022989332675933838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,2560,0.034007999632093645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,1536,0.017775111728244357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,5120,0.025311110748185053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,2048,0.028943111499150593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,1024,0.013653332988421122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,1536,0.023584889041052923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,3072,0.03391022152370877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,768,0.011384000380833944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,1024,0.011393778026103973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,1024,0.01828888886504703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,6144,0.06698311037487455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,512,0.00868888861603207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,768,0.010823999842007955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,2560,0.01682133310370975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,768,0.015189333094490899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,512,0.013190222283204397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,128,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,256,0.011740444435013665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,1536,0.013340444200568728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,64,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,128,0.01145066652033064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,32,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,3584,0.04424533247947693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,64,0.011189333266682096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,256,0.006981333096822103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,512,0.010107555323176915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,65536,0.45816177792019314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,256,0.00958933350112703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,65536,0.24215910169813368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,51200,0.35420089297824436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,128,0.009672000176376766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,51200,0.18905066119299996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,12288,0.101310220029619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,16384,0.060825778378380664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,51200,0.5502595371670193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,16384,0.14804177814059788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,10240,0.09152533610661824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,12288,0.04758844441837735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,8192,0.07889688677257962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,12288,0.11191822422875299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,2048,0.0147706667582194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,10240,0.09620000256432427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,7168,0.05966133541531033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,8192,0.07696266969045003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,16384,0.13665778107113308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,65536,0.6607119772169325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,7168,0.0684337748421563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,6144,0.027457776996824477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,6144,0.059340443876054555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,32,0.011095110740926532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,5120,0.044441776143180005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,10240,0.04123288724157546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,4096,0.03474844495455424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,5120,0.024159999357329473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,3584,0.03352800011634827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,4096,0.04123377799987793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,7168,0.030532442861133154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,6144,0.05915288792716133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,3072,0.02776977750990126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,3584,0.019094222121768527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,3584,0.037147555086347796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,3072,0.017521777086787753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,3072,0.033244444264305964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,2560,0.016113777955373127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,2560,0.02905333373281691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,8192,0.03433510992262099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,2048,0.024492444263564214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,1536,0.017533333765135873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,1536,0.01273155543539259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,1024,0.012855110896958245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,1536,0.020056888461112976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,2560,0.024894222617149353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,768,0.010131555298964182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,1024,0.011069333387745751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,512,0.008016889293988546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,768,0.013120888835854001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,5120,0.050429334243138633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,512,0.009695111049546136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,256,0.006601777755551868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,256,0.009358222285906473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,256,0.010391111175219217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,1024,0.01553244392077128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,128,0.009379555781682333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,2048,0.014392889208263822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,64,0.010093332992659675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,128,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,32,0.0052826665341854095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,2048,0.019887111253208585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,32,0.010073777702119615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,128,0.010048888623714447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,512,0.011718221836619906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,65536,0.24722488721211752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,51200,0.384752008650038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,16384,0.13025244077046713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,51200,0.19358932971954346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,768,0.01036711119943195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,12288,0.09700355264875625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,16384,0.06064177883995903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,16384,0.15600177976820204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,65536,0.505016009012858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,4096,0.020865778128306072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,12288,0.04703288939264086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,8192,0.06382133563359578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,10240,0.041386667225095965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,65536,0.5825786590576172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,10240,0.08556444115108913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,7168,0.055481778250800244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,8192,0.03330399923854404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,6144,0.04830933279461331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,51200,0.4529573122660319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,7168,0.062168002128601074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,6144,0.026941332552168105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,6144,0.05419466561741299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,64,0.00490488898422983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,5120,0.023606222536828782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,5120,0.04665866825315687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,4096,0.019813333948453266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,4096,0.03849777910444472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,8192,0.06999199920230441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,7168,0.030067556434207495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,3072,0.025918222135967676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,3584,0.018411555223994784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,5120,0.04151022103097703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,2560,0.022102221846580505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,3072,0.01585422290696038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,10240,0.07947644260194567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,4096,0.03152533372243246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,2560,0.014369777507252164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,2560,0.026898667216300964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,1536,0.014334221680959066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,2048,0.013847110999955071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,2048,0.02288888891537984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,1024,0.010799110763602786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,1536,0.011403555671374003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,1536,0.01885066595342424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,768,0.008979555633332994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,1024,0.010024888647927178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,3584,0.03480622172355652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,512,0.007335999773608313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,768,0.009646221995353699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,3072,0.03092977735731337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,768,0.012485333614879184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,256,0.006292444550328785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,512,0.009036444127559662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,512,0.010856000085671743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,256,0.008661333057615492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,256,0.009733333355850643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,3584,0.027886221806208294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,64,0.005245333330498801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,128,0.008728000024954477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,32,0.004896000027656555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,1024,0.014496889379289416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,32,0.009402666654851701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,2048,0.018002667360835604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,12288,0.10204622480604385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,65536,0.3646728992462158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,65536,0.20020533932579887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,128,0.009405333134863112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,65536,0.5150044229295518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,16384,0.1032257808579339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,51200,0.15749688943227133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,51200,0.42145599259270566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,64,0.009219555391205682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,16384,0.052213334374957614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,16384,0.12132888370090061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,12288,0.04019022319051955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,12288,0.09268533521228367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,128,0.005247999810510212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,8192,0.054120888312657676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,10240,0.07786133554246691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,51200,0.2769351005554199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,8192,0.02869510981771681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,8192,0.06369066900677152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,7168,0.02706577711635166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,7168,0.05657688776652018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,12288,0.07892711295021905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,6144,0.023324444890022278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,6144,0.04910577668084038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,5120,0.020579554968410067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,5120,0.042295998997158475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,10240,0.03508266806602478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,3584,0.024062222904629175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,4096,0.017504000001483493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,7168,0.04719822274314033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,3072,0.021144888467258875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,3584,0.01624000072479248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,6144,0.039677331844965615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,3584,0.03181155522664388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,3072,0.015119110544522604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,5120,0.029725333054860432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,3072,0.028392887777752344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,2048,0.015435554915004306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,10240,0.0662133362558153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,1536,0.013131555583741931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,2048,0.012492444780137805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,2048,0.02072355482313368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,1024,0.010062221851613786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,1536,0.0173688895172543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,4096,0.035869333479139544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,768,0.008655110994974772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,1024,0.009856888817416297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,1024,0.01350933313369751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,2560,0.018444443742434185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,768,0.009408888717492422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,768,0.011412444214026133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,256,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,512,0.009019555317031013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,2560,0.02459377712673611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,128,0.0052391112678580815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,256,0.008379555410808986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,256,0.009189333352777693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,1536,0.011366222467687396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,64,0.004572444491916233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,32,0.00461777796347936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,128,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,64,0.00868977771864997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,512,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,32,0.00869599978129069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,2560,0.013794666363133324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,65536,0.15776799784766302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,65536,0.4331768883599176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,128,0.008391111261314815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,51200,0.3446933428446452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,65536,0.27990489535861546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,12288,0.06724177466498481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,4096,0.02759644389152527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,12288,0.03227555420663621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,12288,0.07947822411855061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,16384,0.08522222439448039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,10240,0.028621332512961492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,10240,0.06728088855743408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,16384,0.040891554620530866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,8192,0.023183110687467787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,8192,0.05504711137877571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,7168,0.03880177934964498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,7168,0.02106488909986284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,10240,0.05799111392762926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,8192,0.04425599839952257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,6144,0.0425582230091095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,16384,0.10356799761454265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,5120,0.029009779294331867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,6144,0.019069333871205647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,4096,0.023708444502618577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,5120,0.01690933273898231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,5120,0.03677689035733541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,4096,0.01462399959564209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,4096,0.030652443567911785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,51200,0.12733422385321722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,3584,0.013829333086808523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,3072,0.018512000640233357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,3072,0.02437244521247016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,3584,0.0203813330994712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,2560,0.016064888901180692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,6144,0.03290133343802558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,2560,0.011768889096048145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,2560,0.02125155594613817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,2048,0.010882666541470421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,51200,0.26703021261427135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,1536,0.01109777804878023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,3072,0.012808000048001608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,1536,0.009747555686367882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,1024,0.00904177791542477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,1536,0.015104000767072042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,768,0.007305777735180325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,1024,0.0090844440791342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,1024,0.012084444363911947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,2048,0.013348444468445249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,768,0.008662222160233391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,768,0.010404444403118556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,256,0.005196444276306364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,2048,0.01831644442346361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,256,0.007862221863534715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,256,0.008365333080291748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,128,0.004600000050332812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,7168,0.04911999901135763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,64,0.004294222427739037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,128,0.00835111074977451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,3584,0.027688889039887324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,64,0.008177777959240807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,512,0.008104888929261101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,512,0.009347555538018545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,65536,0.22931822141011557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,51200,0.17968532774183485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,65536,0.14411377906799316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,65536,0.3744062317742242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,51200,0.11441511578030056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,512,0.006339555399285422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,51200,0.29789866341484916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,16384,0.03871555460823907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,16384,0.09191466702355279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,32,0.008239111138714684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,10240,0.04693688948949178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,12288,0.03084355592727661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,12288,0.06918222374386258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,8192,0.03703555464744568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,10240,0.02774488925933838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,128,0.007696000238259633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,10240,0.05918311410480075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,7168,0.030781331989500258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,8192,0.022375111778577168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,32,0.004552000098758274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,6144,0.027614222632514104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,7168,0.02125511070092519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,7168,0.043599999613232084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,6144,0.018396443790859647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,6144,0.038218667109807335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,4096,0.019056888090239633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,5120,0.016184000505341422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,5120,0.03304266598489549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,4096,0.014361777239375643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,4096,0.02753688891728719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,3072,0.015220445063379077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,16384,0.07598044474919637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,3584,0.013460444079505073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,2560,0.01350311107105679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,3072,0.012659555508030785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,5120,0.022625777456495497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,3072,0.022030222747060988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,2048,0.011509333219793109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,2560,0.011566222541862063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,2560,0.01927911076280806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,2048,0.010426667001512317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,3584,0.01754488878779941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,1024,0.00794222205877304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,1536,0.009771555662155151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,1536,0.013800889253616333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,768,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,1024,0.009047110875447592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,3584,0.024479111035664875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,1024,0.01124888906876246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,512,0.005723555468850666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,8192,0.048384000857671104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,256,0.004795555439260271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,12288,0.056040002240075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,512,0.008660444782839881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,256,0.007666666474607255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,256,0.007668444679843054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,2048,0.0162471118900511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,64,0.0039013334446483185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,128,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,128,0.00758755538198683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,64,0.007338666253619724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,768,0.008603555460770925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,32,0.007471111085679796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,768,0.009699555734793345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,512,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,65536,0.23353065384758842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,65536,0.13992799652947321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,65536,0.31371378898620605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,51200,0.11111556159125434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,32,0.0042053332759274375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,16384,0.03650310966703627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,16384,0.07958222097820707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,1536,0.009388444324334463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,12288,0.02970577942000495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,51200,0.2601279947492811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,10240,0.03785511189036899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,10240,0.026644445127911035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,12288,0.047648890150917896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,8192,0.04225333200560676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,8192,0.02141688929663764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,7168,0.037512001064088606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,8192,0.03023911184734768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,6144,0.022297778063350256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,7168,0.019322666856977675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,6144,0.017391110459963482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,6144,0.03340177734692892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,7168,0.025050666597154405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,5120,0.015569777952300178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,4096,0.01662400033738878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,10240,0.05141866538259718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,4096,0.013783111340469785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,4096,0.023961777488390606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,3584,0.0151555554734336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,51200,0.17758666144476998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,3072,0.013163555827405719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,3584,0.021526222427686054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,3072,0.012134222520722283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,2560,0.011712888876597086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,2560,0.017072000437312655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,128,0.0042168887125121225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,2560,0.011048888994587792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,5120,0.01980799933274587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,16384,0.06196888950135973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,2048,0.010151999692122141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,2048,0.014775110615624322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,1536,0.009377777576446533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,1536,0.012435555458068848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,12288,0.059917330741882324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,3584,0.012839111189047495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,768,0.005995555470387141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,1536,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,5120,0.02868710954984029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,512,0.005520888914664586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,1024,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,512,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,256,0.004257777912749184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,1024,0.010031999813185798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,2048,0.010130667024188572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,128,0.0041777777175108595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,256,0.007057777709431118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,3072,0.019270223047998216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,128,0.007635555333561367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,768,0.008662222160233391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,32,0.0039022221333450745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,64,0.006735111276308696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,512,0.007829333345095316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,32,0.0069875551594628235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,65536,0.19028000036875406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,1024,0.00867733359336853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,65536,0.2305386596255832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,64,0.003552000141806073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,128,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,16384,0.047192000680499606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,51200,0.18334933121999106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,12288,0.03397333290841844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,16384,0.025820444027582806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,16384,0.05983199675877889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,12288,0.021162667208247717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,12288,0.045928888850741915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,65536,0.0940328902668423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,8192,0.025105777713987563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,10240,0.019544000426928203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,768,0.008375999828179678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,51200,0.07558044460084704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,7168,0.02167466613981459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,8192,0.03258666727277968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,6144,0.01923111081123352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,7168,0.01442577772670322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,7168,0.02943999899758233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,5120,0.01651733285850949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,10240,0.029494222667482164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,6144,0.02612888813018799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,4096,0.014799111419253878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,256,0.007666666474607255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,5120,0.022064000368118286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,10240,0.03900622328122457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,3584,0.01276622215906779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,4096,0.01862133377128177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,8192,0.015680889288584392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,3584,0.01036088913679123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,3584,0.017145777742067974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,2560,0.010064889159467485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,3072,0.009921777579519484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,3072,0.015589333242840238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,6144,0.013173333472675748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,2048,0.008998221821255153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,2560,0.013471999930010902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,51200,0.1448791159523858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,1536,0.007677333222495184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,4096,0.010833777487277985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,1536,0.00832177781396442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,1536,0.010378667049937779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,1024,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,1024,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,768,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,2560,0.009451555709044138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,768,0.007654222349325816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,512,0.004662222332424588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,2048,0.00902844468752543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,2048,0.011900444825490316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,256,0.00406577769252989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,512,0.007263999846246507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,3072,0.011407111254003314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,128,0.003755555384688907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,256,0.006989333364698622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,256,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,64,0.003592888928121991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,1024,0.008671999805503422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,128,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,32,0.003546666767862108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,5120,0.01181777815024058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,65536,0.1394008927875095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,768,0.0075839997993575195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,32,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,51200,0.10916622479756673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,16384,0.03670755690998501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,512,0.006980444822046492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,65536,0.22580001089307997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,51200,0.17562399970160592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,16384,0.025244444608688354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,128,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,10240,0.021584888299306233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,12288,0.044681777556737266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,64,0.006126222097211414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,10240,0.019147555033365887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,8192,0.018610666195551556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,10240,0.03782222337192959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,65536,0.09278666973114014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,8192,0.015554666519165039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,8192,0.03147733211517334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,7168,0.01426666643884447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,7168,0.028392000330819026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,16384,0.057724442746904164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,5120,0.013602666556835175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,6144,0.01314133322901196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,6144,0.02479377720091078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,4096,0.010871999793582492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,5120,0.012048000262843238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,5120,0.021328889661365088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,3584,0.010258666343159145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,7168,0.016407110624843173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,4096,0.018371555540296767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,3072,0.00923644420173433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,3584,0.010393777655230628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,12288,0.025711999999152288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,2560,0.00832177781396442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,3072,0.009832888841629028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,3072,0.015186667442321777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,12288,0.020920000142521326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,2560,0.009370666411187913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,2560,0.013343111508422427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,1536,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,2048,0.011760888828171624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,4096,0.010990222295125326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,1024,0.005280888742870755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,51200,0.07511733637915717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,1536,0.010056888891590966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,768,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,1024,0.008003555238246918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,1024,0.008385777473449707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,768,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,768,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,512,0.007279111279381647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,2048,0.00904266701804267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,1536,0.00831288927131229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,256,0.006008000009589725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,128,0.0038853333228164246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,256,0.006981333096822103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,128,0.006923555499977536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,2048,0.007278222176763747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,128,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,256,0.003923555629120933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,32,0.003560888684458203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,64,0.005965333431959152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,65536,0.10514755381478204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,32,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,6144,0.015069334043396844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,65536,0.183624890115526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,512,0.004304888761705822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,64,0.003551111039188173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,16384,0.03125422199567159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,51200,0.1449848943286472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,16384,0.024452444579866197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,16384,0.0486195551024543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,12288,0.03799555699030558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,3584,0.016580444243219163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,12288,0.020337777005301583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,51200,0.07460977633794148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,10240,0.032623999648623996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,10240,0.018883556127548218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,512,0.006650666809744305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,8192,0.01516533394654592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,7168,0.017865777015686035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,65536,0.09291733635796441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,10240,0.023610666394233704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,7168,0.02443466583887736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,12288,0.026926222774717543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,6144,0.012768000364303589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,5120,0.010478222535716163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,51200,0.08235288990868463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,4096,0.010377777947319878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,5120,0.011601777540312873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,5120,0.018975999620225694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,3584,0.009365333451165093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,4096,0.010570666856235929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,4096,0.01588711142539978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,3072,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,6144,0.021594666772418555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,3584,0.014583999911944071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,2560,0.0069573331210348345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,3072,0.009718221922715506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,8192,0.02698488864633772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,2048,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,2560,0.01182577759027481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,6144,0.011963555382357704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,2048,0.008664888640244802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,8192,0.01941511034965515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,1536,0.00907466643386417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,1024,0.00490399988161193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,1536,0.008039111064540016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,3072,0.013292444248994192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,1024,0.007669332954618666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,2560,0.009112888740168678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,1024,0.007663999994595845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,512,0.0038924444880750445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,768,0.007322666545708974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,7168,0.013812444276279874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,768,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,256,0.0035031110876136353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,512,0.006979555719428592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,128,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,256,0.006650666809744305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,256,0.005689777847793367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,64,0.0032622222271230486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,128,0.005583999885453119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,32,0.0032017777363459268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,768,0.004589333302444882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,64,0.0052879999081293745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,32,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,1536,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,51200,0.04378133349948459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,65536,0.14094577895270452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,3584,0.010053333308961656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,16384,0.017997332745128207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,51200,0.0742408898141649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,51200,0.1132240030500624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,12288,0.015505777464972602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,16384,0.038226667377683855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,65536,0.05161244339413113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,10240,0.014188443620999655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,12288,0.02001066671477424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,12288,0.02994577752219306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,2048,0.010445333189434476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,8192,0.013497778111033969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,10240,0.018599111172888014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,512,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,7168,0.01202311118443807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,8192,0.01499111122555203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,6144,0.010426667001512317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,7168,0.013823111024167804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,7168,0.019697777099079557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,6144,0.012794666820102267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,6144,0.01755288905567593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,5120,0.011501332951916588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,4096,0.008278222547637092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,65536,0.09202933311462402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,10240,0.02569244470861223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,3584,0.007638221813572778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,8192,0.02179466684659322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,4096,0.010388444695207808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,128,0.006640000061856375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,3072,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,3584,0.010109333528412713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,5120,0.009338666995366415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,3584,0.012440889245933957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,3072,0.01145155562294854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,2048,0.005495999836259418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,5120,0.015513777732849121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,2048,0.008635555704434713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,1536,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,4096,0.013377777404255338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,1536,0.008278222547637092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,1024,0.0042088888585567474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,1536,0.00832177781396442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,3072,0.009576888548003303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,2560,0.009026666482289633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,1024,0.00701155596309238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,2560,0.010288889209429422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,768,0.007332444190979004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,768,0.006304889089531369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,512,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,512,0.006953777538405524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,1024,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,768,0.0038897775941424896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,256,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,2048,0.00903555585278405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,512,0.003527111063400904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,64,0.002862222285734283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,128,0.006612444503439798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,128,0.004964444372389051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,16384,0.023631110787391663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,64,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,65536,0.04337244563632541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,256,0.0032666667054096856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,51200,0.03541510966089036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,65536,0.09216888745625813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,65536,0.14015200403001574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,256,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,51200,0.07381333245171441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,51200,0.1128115521536933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,12288,0.010031110710567897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,16384,0.023461333579487268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,32,0.003216888962520493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,2560,0.006301333506902059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,12288,0.01997422178586324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,12288,0.029149333635965984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,32,0.005245333330498801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,10240,0.025397333833906386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,8192,0.014836443795098199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,16384,0.011763555308183035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,7168,0.010418666733635796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,7168,0.013806222213639153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,16384,0.03746666510899862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,10240,0.011746666497654386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,6144,0.01278133359220293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,8192,0.01072799993885888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,6144,0.017503110898865592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,5120,0.01148622234662374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,128,0.0029084444459941653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,8192,0.02123911182085673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,4096,0.01038577738735411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,3584,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,4096,0.01312444441848331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,3584,0.010086221827401055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,3072,0.006607111129495833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,3584,0.012151111331250934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,5120,0.009354666703277165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,2560,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,3072,0.009702222214804756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,4096,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,5120,0.015275556180212231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,2560,0.009019555317031013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,2560,0.010248888697889116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,1536,0.004562666846646203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,7168,0.019270223047998216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,1536,0.008286221987671321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,1536,0.008233778178691864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,10240,0.018639999959203932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,1024,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,768,0.003576888806290097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,3072,0.01125955581665039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,2048,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,512,0.003552000141806073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,768,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,2048,0.008691555923885768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,256,0.003294222263826264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,512,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,512,0.005965333431959152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,128,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,1024,0.003926222109132343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,256,0.005412444472312927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,64,0.0028808888875775864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,128,0.006649777707126405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,128,0.0052062223354975385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,1024,0.0069520001610120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,65536,0.03839200072818332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,64,0.004957333207130432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,51200,0.03151822090148926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,32,0.005160888863934411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,16384,0.011360888679822287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,65536,0.14068355825212267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,256,0.00666933341158761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,16384,0.03708088729116652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,10240,0.009728888670603434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,32,0.0028560000161329904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,8192,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,10240,0.025236444340811834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,6144,0.010407999985747868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,51200,0.10981511407428318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,6144,0.00872444444232517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,7168,0.019348444210158456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,5120,0.008037333687146505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,12288,0.02884355518552992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,4096,0.007654222349325816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,5120,0.015498666299713982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,768,0.006339555399285422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,3584,0.007010666860474481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,4096,0.013250666360060373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,3072,0.006618666566080517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,3584,0.012239110966523489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,2560,0.005631111148330901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,3072,0.011411555939250521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,6144,0.01723911033736335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,2560,0.010307555397351583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,2048,0.009342222577995723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,12288,0.010384000009960597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,1024,0.0038977778620190094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,1536,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,768,0.003903999924659729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,1024,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,512,0.0032115555885765287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,7168,0.009487110707494948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,512,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,128,0.0028364445186323593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,256,0.005484444399674733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,64,0.002557333351837264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,128,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,32,0.002536000063021978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,2048,0.0053119998839166425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,768,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,64,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,51200,0.03148266673088074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,8192,0.021133333444595337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,16384,0.010761777559916178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,65536,0.14288800292544895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,12288,0.013106666505336761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,16384,0.0372284452120463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,10240,0.011796444654464722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,12288,0.02882666720284356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,8192,0.01038044442733129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,65536,0.04373155699835884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,7168,0.009749333063761393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,10240,0.02514222264289856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,1536,0.004597333156400257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,8192,0.021007999777793884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,256,0.0029306666304667792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,5120,0.012436444560686747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,4096,0.008780444661776224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,6144,0.017417778571446735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,5120,0.015451555450757345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,4096,0.013335111240545908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,3072,0.007118222614129384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,3584,0.0123857781291008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,2560,0.006579555571079254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,32,0.004961777892377642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,2048,0.0059262220230367445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,6144,0.014136888914638095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,1536,0.004551110996140374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,7168,0.019333332777023315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,51200,0.1122079955206977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,2048,0.009019555317031013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,768,0.0035333335399627686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,3584,0.008641777767075432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,512,0.003240888938307762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,1024,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,256,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,768,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,128,0.0028435554769304064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,256,0.005281777845488654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,64,0.0028782222006056043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,128,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,2560,0.010058666268984476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,32,0.002519999941190084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,64,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,32,0.004943111290534337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,512,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,3072,0.011126221881972419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,1024,0.003915555361244413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,1536,0.008026666939258575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,2048,0.00904088881280687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,16384,2.1658666398790145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,16384,0.6117342313130697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,16384,2.1313003963894315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,51200,2.1720701853434243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,12288,1.5447777642144098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,12288,0.46924177805582684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,16384,1.8536283704969618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,12288,1.5690284305148656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,10240,1.203736941019694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,10240,0.4031137890285916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,51200,6.324272155761719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,12288,1.3466275533040364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,10240,1.2753582000732422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,8192,0.3167360093858507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,8192,0.9665875964694552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,8192,1.0170461866590712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,7168,0.798799991607666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,10240,1.1016666624281142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,7168,0.2924986680348714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,51200,8.18546379937066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,7168,0.8438755671183268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,6144,0.6348746617635092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,8192,0.8585804303487142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,6144,0.7537733183966743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,7168,0.768830246395535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,5120,0.6109786563449436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,5120,0.6096595658196343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,5120,0.21679911348554823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,6144,0.659530692630344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,4096,0.4924631118774414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,5120,0.5397760073343912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,6144,0.24757512410481772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,4096,0.17945067087809244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,3584,0.40409689479404026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,4096,0.47379642062717015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,3584,0.1605831119749281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,3584,0.4366017712487115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,4096,0.4556879997253418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,3072,0.3378400007883708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,3584,0.38721688588460285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,2560,0.31822845670912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,3072,0.34781866603427464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,2560,0.31027200486924916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,2048,0.22937954796685112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,51200,5.9922434488932295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,2048,0.25127466519673664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,2048,0.10348711411158244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,3072,0.35832799805535215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,2048,0.24307110574510363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,1536,0.1991342173682319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,2560,0.14738578266567653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,1024,0.1345004505581326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,2560,0.28560622533162433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,1024,0.06697333521313138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,768,0.10881333218680488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,1024,0.14973066912757024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,1024,0.14149155881669787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,1536,0.20264444086286756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,512,0.08541955550511678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,3072,0.14062133100297716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,768,0.12149511443244086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,768,0.11513245105743408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,512,0.052975111537509494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,512,0.104839113023546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,1536,0.18529510498046875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,128,0.04415200153986613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,256,0.09377066956626044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,256,0.049066666099760264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,256,0.07541955841912164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,64,0.04044977823893229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,128,0.046801779005262584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,128,0.08751110898123847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,768,0.05905244747797648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,128,0.05929244226879544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,64,0.08728444576263428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,32,0.08796799845165676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,512,0.09477600124147202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,32,0.03978488842646281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,256,0.051538666089375816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,1536,0.0849377777841356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,65536,2.2424916161431208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,51200,1.7330480151706273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,51200,4.953252580430773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,16384,1.8259752061631944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,16384,1.7538400226169164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,65536,6.569591946072048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,16384,0.5266995429992676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,12288,1.198911984761556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,12288,1.1674817403157551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,51200,6.003169589572483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,12288,0.3701653215620253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,65536,8.669798109266493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,16384,1.4009990692138672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,10240,0.3086186779869927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,12288,1.0253164503309462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,8192,0.7501617537604438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,8192,0.2657617727915446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,8192,0.7995297643873426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,10240,0.8756426705254449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,10240,0.9122373792860242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,51200,4.710012647840712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,8192,0.681487136416965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,7168,0.6946658028496636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,7168,0.2242586612701416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,6144,0.5446631113688151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,7168,0.6667937702602811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,6144,0.19651822249094644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,5120,0.4475368923611111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,7168,0.6074942482842339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,5120,0.47426488664415145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,10240,0.9580675760904948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,6144,0.5206969049241807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,4096,0.3927786615159776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,4096,0.14401422606574163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,3584,0.3308275540669759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,4096,0.3762639893425836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,4096,0.3423146671719021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,65536,6.148575252956814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,3584,0.1254835526148478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,3584,0.3263804382748074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,3072,0.24868443277147081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,3584,0.31483644909328884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,3072,0.11527289284600152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,3072,0.28311289681328666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,5120,0.4240399996439616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,2560,0.24089956283569336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,2560,0.0972977810435825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,2560,0.2349475489722358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,2048,0.20518578423394096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,2048,0.08421422375573052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,5120,0.17012977600097656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,2560,0.20922577381134033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,1536,0.15956711769104004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,1536,0.06690400176578097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,1536,0.1575662162568834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,3072,0.2659413284725613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,1024,0.11588888698154026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,6144,0.6002248658074273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,1536,0.14800000190734863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,768,0.08071555693944295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,1024,0.05344533258014255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,1024,0.11861956119537354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,512,0.057816889550950795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,768,0.04780888888570997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,2048,0.19349066416422525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,768,0.09684711032443577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,256,0.042048000627093844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,512,0.04324266645643446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,512,0.07610399855507745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,256,0.07473688655429415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,128,0.035270220703548856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,256,0.06086933612823486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,1024,0.11208266682094997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,2048,0.19828355312347412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,64,0.031010667483011883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,128,0.06969955232408312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,32,0.03170933326085409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,128,0.048455999957190625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,64,0.0696915586789449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,512,0.08300355407926771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,32,0.07024088833067152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,256,0.04005511270629035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,128,0.038175112671322294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,768,0.09242577685250176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,65536,0.699127991994222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,51200,1.5697013007269964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,65536,2.154971652560764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,51200,0.5403395758734809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,16384,0.6029991043938531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,65536,2.400517357720269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,16384,0.16253244876861572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,12288,0.41655733850267196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,16384,0.5219866434733073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,12288,0.127527117729187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,16384,0.4396346675025092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,12288,0.3286666605207655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,10240,0.10698755582173665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,10240,0.2769075499640571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,51200,1.5750399695502386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,8192,0.2766604423522949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,12288,0.37471821573045516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,8192,0.2518257829878065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,8192,0.2322782145606147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,10240,0.3144017855326335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,7168,0.2191013362672594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,51200,2.0286142561170792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,7168,0.07999910910924275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,10240,0.32280177540249294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,6144,0.1876986689037747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,7168,0.19438221719529894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,6144,0.06914044751061334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,65536,2.0032791561550565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,7168,0.20336711406707764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,5120,0.14504355854458278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,4096,0.13692088921864828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,8192,0.09214578072230022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,4096,0.12654488616519505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,5120,0.17003999816046822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,4096,0.1216240061653985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,3584,0.11151466104719375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,6144,0.21137689219580758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,3584,0.10785422060224746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,3072,0.10368000136481391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,3072,0.0976311100853814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,4096,0.049568000766966075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,3072,0.09427022271686131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,2560,0.08316622177759807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,3072,0.04054044352637397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,5120,0.1575582159890069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,2560,0.08424178096983169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,2048,0.06159733401404487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,2560,0.08221777942445543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,3584,0.045365333557128906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,1536,0.049301332897610135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,2048,0.030626667870415583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,2048,0.06992710961235894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,6144,0.17465955681271025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,2048,0.07111910978953044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,3584,0.11952800220913357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,1536,0.056905779573652476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,1024,0.042865776353412204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,1024,0.020717332760492962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,768,0.025900445050663416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,1024,0.04429422153366936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,5120,0.05934844414393107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,768,0.03403733174006144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,768,0.018625777628686693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,512,0.01979733341270023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,1024,0.03606399893760681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,1536,0.024653333756658766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,256,0.015934222274356417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,512,0.016879111528396606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,512,0.03177422285079956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,2560,0.03590755661328634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,128,0.017885333961910672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,256,0.01606044504377577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,256,0.025027554896142747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,128,0.02477333280775282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,128,0.015564444992277356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,64,0.015042665931913586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,128,0.021643555826610986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,32,0.01555199921131134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,1536,0.055474665429857045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,512,0.028583112690183852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,64,0.02441155579355028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,32,0.02443644404411316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,256,0.02583644456333584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,768,0.03785689009560479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,51200,1.0424942440456815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,65536,1.519926177130805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,16384,0.33449334568447536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,51200,0.39913156297471786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,65536,1.7925804985894098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,51200,1.4765564600626628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,16384,0.13702489270104304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,65536,0.5125475459628636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,16384,0.3380880090925429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,51200,1.1084470748901367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,10240,0.24509421984354654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,65536,1.4595040215386286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,10240,0.08295200268427531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,10240,0.2742951181199816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,16384,0.39443378978305393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,12288,0.2644888824886746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,10240,0.21739644474453398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,7168,0.17007466157277426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,8192,0.19306311342451307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,12288,0.2936462296379937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,8192,0.17585333188374838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,12288,0.2479475604163276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,6144,0.14945867326524523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,7168,0.1683128939734565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,7168,0.15026488569047716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,6144,0.05408089028464424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,5120,0.11886755625406902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,6144,0.14312711026933458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,6144,0.1343617836634318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,4096,0.10084444284439087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,12288,0.09721155299080743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,5120,0.12205333179897732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,5120,0.11248266696929932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,4096,0.03838133480813768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,7168,0.05973155630959404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,4096,0.09717688957850139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,4096,0.0925111108356052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,3584,0.03443911009364658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,3584,0.08663022518157959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,8192,0.18839111593034533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,3072,0.07601955864164564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,3072,0.07386222150590685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,3584,0.08514222171571519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,2560,0.06565688716040717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,3072,0.07404088973999023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,2048,0.05281866590181986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,3584,0.08282044198777941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,3072,0.030741333961486816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,2048,0.05468089050716824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,2048,0.02424266603257921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,1536,0.044180442889531456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,2560,0.027845333019892376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,1536,0.02010577751530541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,5120,0.04563022322124905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,2560,0.06757955418692695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,1536,0.035114665826161705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,1024,0.03405333227581448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,768,0.020486222373114694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,2560,0.06396355231602986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,768,0.027512888113657635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,8192,0.06831555234061347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,512,0.016050666570663452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,2048,0.05463644531038073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,1024,0.016993777619467843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,1024,0.03530133432812161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,256,0.012072000238630505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,512,0.014126222994592456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,256,0.020231111182106864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,256,0.019918221566412184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,128,0.010885333021481832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,128,0.019824888971116807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,512,0.022706665926509436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,512,0.025635555386543274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,64,0.009692444569534725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,128,0.013454222016864352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,32,0.009719111025333405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,128,0.017504889104101393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,1536,0.04495644569396973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,256,0.013234666652149625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,32,0.019297776950730216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,65536,1.2075555589463975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,1024,0.026421333352724712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,768,0.029959112405776978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,65536,1.6709733539157445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,65536,0.45770221286349827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,51200,0.9205795923868815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,768,0.015375110838148328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,16384,0.32841867870754665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,51200,1.2176390753851998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,51200,0.3569840060340033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,16384,0.11014578077528213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,12288,0.26646576987372506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,16384,0.294140444861518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,12288,0.08466222551133896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,10240,0.2213644451565213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,12288,0.21697689427269828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,65536,1.249747594197591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,10240,0.20124267207251656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,8192,0.17571466498904756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,10240,0.18476711379157174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,8192,0.1591591172748142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,8192,0.059438221984439425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,12288,0.23566754659016928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,7168,0.15290666951073542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,64,0.019654222660594516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,51200,0.9391031265258789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,7168,0.05268266797065735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,7168,0.14250044027964273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,7168,0.13109066751268175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,10240,0.07161421908272637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,6144,0.04654933346642388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,6144,0.12242044342888726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,16384,0.31826310687594944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,6144,0.11440977785322402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,4096,0.0837271081076728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,5120,0.04034488730960422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,8192,0.14834400018056235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,5120,0.0986586676703559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,3584,0.07690311140484281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,4096,0.08339022265540229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,4096,0.08332266410191853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,6144,0.12979288895924887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,3584,0.030312889152103003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,3584,0.07328177822960748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,5120,0.10897688733206855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,2560,0.054656889703538686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,3072,0.06515022118886311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,5120,0.10335644748475815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,2560,0.05575022432539198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,2048,0.04541600081655714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,3072,0.06586488750245836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,2048,0.04666044314702352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,3072,0.06521422333187528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,2048,0.021409778131379023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,2048,0.04941511154174805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,4096,0.03328533305062188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,1536,0.03770666652255588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,2560,0.05704266495174832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,1536,0.017832888497246634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,3584,0.07440711392296685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,1024,0.029304888513353135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,768,0.020940444535679285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,1024,0.03182399935192532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,1536,0.0347013341055976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,768,0.014134221606784396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,2560,0.024703111913469102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,1024,0.025591111845440332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,512,0.019828443725903828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,1024,0.015061333775520325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,512,0.022777777579095628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,512,0.012887111140622033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,3072,0.02737955583466424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,256,0.012135110795497894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,256,0.017626666360431247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,768,0.023757333556811016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,256,0.01874755488501655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,512,0.018040888839297824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,64,0.011748444702890186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,128,0.012133333418104382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,128,0.017344888713624742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,128,0.01713244451416863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,64,0.01686044368478987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,32,0.016904888881577384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,128,0.009677333136399588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,768,0.026888888743188646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,65536,0.9446257485283746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,32,0.01178311142656538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,65536,0.372104008992513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,256,0.014420444766680399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,51200,0.87500090069241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,16384,0.27350489298502606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,51200,0.29103732109069824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,51200,0.9070542653401693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,1536,0.04071377714474996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,65536,1.039732403225369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,16384,0.2802177800072564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,16384,0.2554639975229899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,51200,0.8180301984151205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,12288,0.19326489501529268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,12288,0.0717031094763014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,10240,0.16529333591461182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,12288,0.18942666053771973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,10240,0.16264800230662027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,10240,0.0612915555636088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,16384,0.09133866760465835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,12288,0.1797999938329061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,10240,0.15742222468058267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,8192,0.04941244588957893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,8192,0.1292711098988851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,7168,0.11711022588941787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,7168,0.04373688830269707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,6144,0.08669422070185344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,7168,0.11445600456661648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,8192,0.13565866152445474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,6144,0.03837955660290188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,5120,0.08423822455936009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,6144,0.09946578078799778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,7168,0.11725778049892849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,5120,0.08513333400090535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,5120,0.033597333563698664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,65536,1.2748497856987846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,5120,0.08591466479831272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,3584,0.05824977821773953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,4096,0.02879377868440416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,4096,0.068996442688836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,3072,0.050520890288882785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,3584,0.02625333269437154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,3584,0.06468977530797322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,6144,0.10110044479370117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,2560,0.0461422238085005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,8192,0.13140088982052273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,4096,0.06637510988447401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,2560,0.04652977652019925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,2560,0.02197511163022783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,2560,0.051035556528303355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,4096,0.07113866673575507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,2048,0.03944266504711575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,1536,0.026612444056404963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,3072,0.024030221833123103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,3072,0.05730488565233019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,1536,0.016509332590632968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,1024,0.019694222344292533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,2048,0.03480089041921828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,3584,0.061426666047838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,1024,0.024716445141368445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,1024,0.014053333136770459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,2048,0.018868444694413077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,768,0.020071999894248117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,1536,0.03224444389343262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,512,0.01241422196229299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,768,0.012984000146389008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,1536,0.035269333256615534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,512,0.011798222031858234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,512,0.016920889417330425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,256,0.012071111136012606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,512,0.020624889267815482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,256,0.015224888920783997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,256,0.016472889317406546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,768,0.023760888311598036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,128,0.014667555689811707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,128,0.011389333340856763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,64,0.00963644435008367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,128,0.015535111228624979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,768,0.016200888488027785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,32,0.009682666924264696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,64,0.014536889062987434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,32,0.01447111037042406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,65536,0.8881520165337456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,1024,0.02801333202256097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,256,0.011113777756690979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,65536,0.33088710572984487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,51200,0.7654142379760742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,65536,0.9054088592529297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,16384,0.2643582291073269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,51200,0.2537413438161214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,128,0.010510221951537661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,51200,0.6737910906473795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,16384,0.0801688896285163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,16384,0.2156640026304457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,12288,0.17179555363125273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,2048,0.043671111265818276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,51200,0.8438124656677246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,12288,0.16181243790520564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,10240,0.14344444539811876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,65536,1.103384017944336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,8192,0.12703911463419595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,16384,0.23085689544677734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,10240,0.14430756039089626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,10240,0.13599644766913518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,8192,0.044312887721591525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,8192,0.11796443992190891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,8192,0.11314311292436387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,3072,0.05398222141795688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,7168,0.039100444979137845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,12288,0.062353776560889355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,7168,0.09983911116917928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,10240,0.053971555497911244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,6144,0.0898195571369595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,6144,0.08808710840013291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,7168,0.10835466782251994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,5120,0.07620355818006727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,6144,0.09420888953738743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,7168,0.10371288988325332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,5120,0.030233777231640283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,4096,0.06225866741604275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,12288,0.20423377884758842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,3584,0.058711998992496066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,3584,0.05560355716281467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,5120,0.07524889045291476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,3584,0.023608888188997906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,3584,0.05640711386998495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,3072,0.04974400003751119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,6144,0.03514488869243198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,4096,0.025796444879637823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,3072,0.02169688873820835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,3072,0.04889511068662008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,3072,0.05076622300677829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,2048,0.03427733315361871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,2560,0.01959466603067186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,4096,0.06588177548514472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,2048,0.017077333397335477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,1536,0.02664888898531596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,5120,0.08316444688373141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,2560,0.04277599851290385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,1536,0.0295466681321462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,1536,0.014764444695578681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,1024,0.020050666398472257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,1536,0.03145688772201538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,2560,0.042416890462239586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,768,0.017121778594122995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,1024,0.012598222328556908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,1024,0.022513777017593384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,1024,0.025399999486075506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,2048,0.03569599986076355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,768,0.011748444702890186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,768,0.018553778529167175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,4096,0.06341422266430326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,256,0.00923111124171151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,512,0.01110577748881446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,512,0.015691555208630033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,512,0.0184515549076928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,128,0.008346666892369589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,256,0.010671111444632212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,256,0.01389600005414751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,256,0.014879999889267815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,64,0.0069955554273393415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,128,0.013463111387358772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,32,0.00905688852071762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,512,0.011863110793961419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,2048,0.03768710957633124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,64,0.013464888764752282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,768,0.021399110555648804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,32,0.013473778135246701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,65536,0.6880790922376844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,65536,0.2818400065104167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,51200,0.4677128791809082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,2560,0.045258667733934194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,128,0.01517955627706316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,65536,0.9154657787746854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,51200,0.21346399519178602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,65536,0.7219973140292697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,16384,0.20387910472022164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,12288,0.14111822181277806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,51200,0.5694684452480739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,128,0.010443555812040964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,12288,0.15323111746046278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,12288,0.05623200204637316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,10240,0.1206551127963596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,12288,0.13866489463382298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,8192,0.09816977712843154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,10240,0.04591022266281975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,10240,0.12815645005967882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,10240,0.1177342202928331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,16384,0.06783644358317058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,8192,0.03652711047066583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,7168,0.08819111188252766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,8192,0.09551555580563015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,7168,0.03319022059440613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,7168,0.09193066755930583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,7168,0.08487911356820001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,16384,0.1577155590057373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,5120,0.05191466543409559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,6144,0.029793779055277508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,6144,0.07408266597323947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,51200,0.7573999828762479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,5120,0.06777777936723497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,5120,0.06416889031728108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,6144,0.07318133115768433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,4096,0.05574755536185371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,4096,0.023045332895384893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,6144,0.07880000273386638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,3584,0.04945333467589485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,5120,0.026141333911154006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,4096,0.04965155654483371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,3072,0.03851822349760268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,3072,0.04354666670163473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,4096,0.05419644382264879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,3072,0.04348266786999173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,2560,0.03100444542037116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,3072,0.01907111042075687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,2560,0.037625776396857366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,2560,0.03812177644835578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,2048,0.025519111090236243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,3584,0.043269332912233144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,2048,0.031825777557161115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,2048,0.015503111812803479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,1536,0.019651555352740817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,2048,0.033654222885767617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,3584,0.04901688959863451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,1536,0.026231999198595684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,1024,0.01462933255566491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,1024,0.01982399986849891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,3584,0.02128355536195967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,1024,0.011436444189813403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,768,0.012411555482281579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,2560,0.01724799970785777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,1536,0.02806666824552748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,512,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,768,0.010785777535703448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,768,0.018962666392326355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,512,0.014150222142537435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,512,0.010103111465771994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,256,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,512,0.01641244524055057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,256,0.009682666924264696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,256,0.01277955538696713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,256,0.013740444348918067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,1536,0.013439111411571503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,128,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,128,0.01240355521440506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,64,0.005969777703285217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,32,0.0063635553750726916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,64,0.012154666913880242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,32,0.012068444656001197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,16384,0.1808782286114163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,128,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,65536,0.562380419837104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,768,0.016640888320075143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,65536,0.25296089384290904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,128,0.013439999686347114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,51200,0.4612693256802029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,65536,0.7810284296671549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,16384,0.15742222468058267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,51200,0.19908889134724936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,8192,0.10229777627521092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,51200,0.5946995417277018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,65536,0.6920968691507975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,16384,0.06343111064698961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,51200,0.5426764488220215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,12288,0.0506640010409885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,16384,0.17635111014048258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,12288,0.12940355141957602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,10240,0.1104017761018541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,12288,0.13321867254045275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,8192,0.0895306666692098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,10240,0.04336177640491062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,10240,0.10959111319647895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,10240,0.11248355441623265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,8192,0.03502488798565335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,8192,0.0886568890677558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,1024,0.022101332743962605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,8192,0.09521866507000393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,16384,0.1721715529759725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,7168,0.03171644608179728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,7168,0.07829688655005561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,7168,0.08244533008999295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,6144,0.028018666638268366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,6144,0.06826221942901611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,6144,0.07234844234254625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,5120,0.025111999776628282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,5120,0.05805244710710314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,5120,0.06204533576965332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,7168,0.07998310857348971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,3584,0.040060444010628596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,4096,0.04752444558673435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,6144,0.06490666336483426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,4096,0.05238222082455953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,3072,0.03401688734690348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,3584,0.019976889093716938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,5120,0.054904000626669995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,3584,0.04746133420202467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,3072,0.018394667241308425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,4096,0.04603022336959839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,3072,0.04234310984611511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,2560,0.016399999459584553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,2048,0.025158221522967022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,2560,0.03700266612900628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,12288,0.12377688619825576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,2048,0.01459822224246131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,1536,0.02012977831893497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,2048,0.03215377860599094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,2560,0.028839998775058325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,1536,0.013043555948469373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,1024,0.01551644504070282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,2560,0.03280444608794318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,4096,0.02164977788925171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,1024,0.017502221796247695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,3584,0.04298222064971924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,1024,0.011380444798204633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,3072,0.03756711218092177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,768,0.014867555763986377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,1536,0.02680533296532101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,512,0.009684444301658207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,512,0.012607999973826937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,1024,0.021035555336210463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,768,0.013445333474212222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,256,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,512,0.010062221851613786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,512,0.016181333197487723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,1536,0.022535110513369244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,256,0.011416888899273343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,256,0.009373333719041612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,256,0.013048888908492194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,128,0.009390222529570261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,128,0.01107377807299296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,128,0.012159999873903064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,2048,0.027874665127860174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,32,0.006796444455782573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,64,0.011078221930397881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,32,0.011000889043013254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,128,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,768,0.018231999542978074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,65536,0.204821334944831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,64,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,65536,0.6284373071458604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,16384,0.12564622031317815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,51200,0.1617235607571072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,51200,0.4820959832933214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,51200,0.418579551908705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,16384,0.052170667383405894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,65536,0.5287137561374241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,12288,0.10023822387059529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,768,0.010647111468844943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,16384,0.13878933588663736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,10240,0.09363910886976455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,12288,0.042711999681260854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,12288,0.10871377918455337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,12288,0.10679288705190022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,10240,0.03666222095489502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,65536,0.5350026554531521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,10240,0.09072888559765285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,8192,0.029511110650168523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,7168,0.06442755460739136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,8192,0.07600622044669257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,16384,0.14482933945126003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,6144,0.049752000305387706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,7168,0.026198221577538386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,7168,0.06739999850591023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,51200,0.3926435576544867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,5120,0.04894666539298164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,10240,0.09188710980945164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,6144,0.05950044261084663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,5120,0.021235555410385132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,5120,0.04960888955328199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,4096,0.037485334608289934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,7168,0.06743822495142619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,4096,0.041131556034088135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,3584,0.03264800045225356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,4096,0.043303112188975014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,8192,0.07714311281840007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,3584,0.017083555459976196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,3072,0.030026667647891577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,8192,0.07450755437215169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,5120,0.051494220892588295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,3072,0.03306844499376085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,3072,0.03504088852140639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,4096,0.01864177816443973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,6144,0.023825777901543513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,2560,0.028780443800820246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,2048,0.020604444874657523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,3584,0.03910488883654276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,2560,0.03158933255407546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,2048,0.013291555146376291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,1536,0.01482488877243466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,3072,0.015722667177518208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,2560,0.0244168887535731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,1536,0.020059555768966675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,6144,0.05814133087793986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,1536,0.022156443860795762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,3584,0.03704799877272712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,1024,0.015275556180212231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,1024,0.010111110905806223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,1024,0.018049778209792245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,2048,0.02719288898838891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,768,0.013146667016877068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,512,0.00797333319981893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,768,0.009687999884287516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,768,0.015860444969601102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,1024,0.010853333605660332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,256,0.007012444237867991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,512,0.00904622260067198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,512,0.01146399974822998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,512,0.014162666267818876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,128,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,768,0.009373333719041612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,256,0.010103999740547603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,2048,0.02421155571937561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,128,0.009696000152164036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,64,0.007009777757856581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,1536,0.01146488885084788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,32,0.007191999918884701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,64,0.009688888986905416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,32,0.009722666607962714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,256,0.012303999728626676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,65536,0.47226042217678493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,128,0.008660444782839881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,128,0.012093333734406365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,65536,0.19677066802978516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,51200,0.3804924488067627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,2560,0.014537778165605334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,65536,0.562447124057346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,16384,0.11681600411732991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,256,0.008820444345474243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,65536,0.5290071169535319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,51200,0.1585199965371026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,51200,0.43930578231811523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,12288,0.09371288617451985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,16384,0.05039377676116096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,16384,0.13012266159057617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,10240,0.08798933029174805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,12288,0.04042844308747186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,12288,0.09925688637627496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,12288,0.10456977950202094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,8192,0.06978933016459148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,10240,0.08432977729373509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,10240,0.08980621894200642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,8192,0.028140445550282795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,7168,0.05924444728427463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,8192,0.06826400094562106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,8192,0.07411911090215047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,6144,0.05199377735455831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,7168,0.025644444757037695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,7168,0.06076444519890679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,6144,0.02334044377009074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,6144,0.053039110369152494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,16384,0.13689333862728542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,6144,0.05806844764285617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,4096,0.034852445125579834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,5120,0.045668443044026695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,5120,0.050373332368003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,4096,0.03769333495034112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,4096,0.018166222506099276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,3584,0.031204445494545832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,4096,0.04273600048489041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,3584,0.016727111405796476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,7168,0.06782844331529406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,3072,0.02713955607679155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,3584,0.038578665918774076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,10240,0.03531644410557217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,3072,0.015419556034935845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,2560,0.02312444481584761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,3072,0.03470933437347412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,2560,0.01422755585776435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,2048,0.019848888119061787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,5120,0.042232887612448804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,2048,0.022784888744354248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,5120,0.02071644365787506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,2048,0.012786666552225748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,1536,0.016048888365427654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,2048,0.02618933386272854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,1536,0.011395555403497485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,2560,0.02648177742958069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,2560,0.030720889568328857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,51200,0.4127511183420817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,1024,0.017610667480362784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,1024,0.010040889183680216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,768,0.009378666679064432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,3072,0.03092977735731337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,768,0.012294222083356647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,512,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,1536,0.01889866590499878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,1536,0.021886222892337378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,512,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,3584,0.03449777762095133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,256,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,256,0.009697777529557547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,256,0.011413333316644033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,128,0.0052933332820733385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,256,0.008555555509196388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,768,0.015600000818570455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,1024,0.011744000017642975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,64,0.005179555465777715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,128,0.00850577818022834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,128,0.009353777600659264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,32,0.005045333256324132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,512,0.010728889041476779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,1024,0.014214222629865011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,512,0.013705777625242868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,32,0.009016888837019602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,65536,0.3672791057162815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,65536,0.16274666786193848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,51200,0.2678568893008762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,768,0.009397333694828881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,65536,0.397669341829088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,16384,0.10097422202428182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,51200,0.38276089562310117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,128,0.011042666931947073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,64,0.00923822240697013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,16384,0.041660444604025945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,12288,0.08100533485412598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,65536,0.4867288801405165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,12288,0.0807466639412774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,51200,0.12621243794759116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,51200,0.3105200131734212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,16384,0.11595822705162896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,10240,0.0742737783326043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,16384,0.10626666413413154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,8192,0.051560001240836255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,10240,0.028322666883468628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,10240,0.06894399722417195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,7168,0.045942223734325834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,12288,0.08681866857740615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,8192,0.06030488676495022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,8192,0.05797955724928114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,12288,0.03200799889034695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,7168,0.05344355437490675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,7168,0.05202310946252611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,6144,0.04684799909591675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,6144,0.019073777728610568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,5120,0.03306488858328925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,6144,0.04587022132343716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,4096,0.027083555857340496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,5120,0.01702222228050232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,5120,0.04009777638647292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,5120,0.039968000517951116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,3584,0.02276977731121911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,10240,0.0633537769317627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,4096,0.03408711155255636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,3584,0.03021777669588725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,3072,0.02020355562369029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,8192,0.02292622294690874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,3072,0.02658222284581926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,3072,0.012798222402731577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,4096,0.033486220571729876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,2560,0.017539555827776592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,3072,0.02764355474048191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,3584,0.014159111513031853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,2560,0.02346666653951009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,2048,0.015092444088723926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,2560,0.02442488902144962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,2048,0.01070666644308302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,2048,0.019662222928471036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,2048,0.021846221552954778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,4096,0.014877332581414117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,6144,0.03921422362327576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,1536,0.016404444972674053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,1536,0.009791110952695211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,1536,0.017668444249365065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,2560,0.011765333513418833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,1024,0.0090515555606948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,1024,0.014813333749771118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,1536,0.012474666866991254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,768,0.011383111278216044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,512,0.006648889018429651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,768,0.008662222160233391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,768,0.01349422252840466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,7168,0.02126666737927331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,1024,0.009384888741705153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,512,0.008343111309740279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,512,0.01183733344078064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,1024,0.013053333593739403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,128,0.004605333424276776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,256,0.010388444695207808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,128,0.008737777670224508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,128,0.01072088877360026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,64,0.004228444563017951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,256,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,32,0.004576000074545543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,64,0.008671110702885522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,768,0.008349333372380998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,256,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,65536,0.34461066457960343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,65536,0.4276106622483995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,512,0.00983911090426975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,51200,0.21483822663625082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,3584,0.030846221579445735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,32,0.008718222379684448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,16384,0.08479644192589654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,51200,0.11965866883595784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,51200,0.300289789835612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,65536,0.1526684496137831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,16384,0.10109155707889134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,16384,0.03901955485343933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,65536,0.38167556126912433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,12288,0.031211555004119873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,256,0.008808888494968414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,10240,0.056392000781165234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,12288,0.0793057746357388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,51200,0.3330426745944553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,10240,0.027570666538344488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,10240,0.06640177965164185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,10240,0.06776711013582018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,128,0.008010666403505538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,8192,0.054063111543655396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,8192,0.022678222921159532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,8192,0.057074666023254395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,12288,0.0777626633644104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,7168,0.048471109734641186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,6144,0.035884443256590105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,7168,0.020946666598320007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,8192,0.0422017772992452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,6144,0.04214400053024292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,6144,0.018722666634453666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,5120,0.029727111260096233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,6144,0.04589688777923584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,7168,0.03680888811747233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,16384,0.10179911057154338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,4096,0.023948444260491267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,5120,0.016703999704784818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,5120,0.03640800052218967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,3584,0.021425777011447485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,4096,0.014840889308187695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,4096,0.030636443032158747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,7168,0.050804442829555936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,4096,0.03352888756328159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,3072,0.019304000669055514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,3584,0.013804444836245643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,3584,0.030945777893066406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,3072,0.01276266657643848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,3072,0.024714666936132643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,3072,0.02721600068940057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,12288,0.06990310880872938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,2048,0.014911111858155994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,2560,0.011581333147154914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,2560,0.02445777753988902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,5120,0.039078222380744085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,1536,0.012096888489193387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,2048,0.010739555789364709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,2048,0.020890666378868952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,3584,0.027699554959932964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,1536,0.015242666006088257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,1536,0.017622222503026325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,2560,0.016951999730534024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,768,0.007867555651399825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,2560,0.021078222327762183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,1024,0.009002666506502364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,768,0.010248888697889116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,512,0.006614222294754452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,2048,0.018002667360835604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,768,0.013221333424250284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,1536,0.009758222434255812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,256,0.005647111270162795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,512,0.009373333719041612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,512,0.011953777737087674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,256,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,256,0.008365333080291748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,128,0.005240889059172736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,256,0.01039644413524204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,768,0.0086986662613021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,128,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,128,0.010384000009960597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,32,0.005633777628342311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,512,0.008264888491895463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,64,0.008011555506123437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,32,0.00796977761718962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,65536,0.2864026758405897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,1024,0.014829332629839579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,65536,0.14779910776350233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,128,0.007691555553012424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,64,0.004576888763242298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,65536,0.3835831218295627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,51200,0.11354845099978977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,16384,0.07439021931754218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,51200,0.2932648923661974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,51200,0.30270398987664116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,16384,0.03724177678426107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,12288,0.053661333190070264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,16384,0.0901342233022054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,1024,0.012141333685980903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,16384,0.10194400284025405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,10240,0.044010665681627065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,12288,0.029992000924216375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,51200,0.20492622587415907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,8192,0.03683288892110189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,12288,0.08046399884753756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,10240,0.058337774541642934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,1024,0.009462222456932068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,10240,0.06696889135572645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,7168,0.03175910976197984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,8192,0.02164622147878011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,8192,0.04831288920508491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,6144,0.027666666441493567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,7168,0.0199306673473782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,7168,0.04306133257018196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,7168,0.04998044504059685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,5120,0.023394667439990576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,6144,0.03782755467626784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,12288,0.06893244716856214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,10240,0.027087110612127516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,5120,0.015971556305885315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,5120,0.03836977812978957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,8192,0.05736622545454237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,4096,0.03296000096533034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,4096,0.014194667339324951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,3584,0.017817777064111497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,6144,0.01834755473666721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,65536,0.3708871205647786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,6144,0.04422577884462145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,3072,0.01604711181587643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,3584,0.013487999637921652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,4096,0.019863110449579027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,3072,0.012415111064910889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,2560,0.01387555566098955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,3072,0.026737777723206416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,2560,0.01924088928434584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,2560,0.02349688940578037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,2048,0.011797332929240333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,3584,0.024784889486100938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,3584,0.029703110456466675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,2048,0.01625155574745602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,2048,0.010400888820489248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,2048,0.020599111914634705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,1536,0.01391999920209249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,1536,0.009712888962692684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,2560,0.011423110961914062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,5120,0.032784889141718544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,1024,0.01108000013563368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,3072,0.02200266718864441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,1024,0.008780444661776224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,4096,0.027151111099455092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,1024,0.0144195556640625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,512,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,768,0.008370666868156856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,768,0.013108444710572561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,1536,0.01720977822939555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,1024,0.007678222325113084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,512,0.008023111356629265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,512,0.012080888781282636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,256,0.007694222033023834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,256,0.007671111159854465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,128,0.004562666846646203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,256,0.010064000056849586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,128,0.007751999629868402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,128,0.007322666545708974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,768,0.009400889277458191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,128,0.010388444695207808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,32,0.004325333154863781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,512,0.008664888640244802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,64,0.007336888876226213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,32,0.007329777710967594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,65536,0.0981280008951823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,65536,0.29241421487596303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,51200,0.17523554960886636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,64,0.004104888687531154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,51200,0.22984088791741264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,51200,0.07816355758243136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,65536,0.22412710719638398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,51200,0.1862177716361152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,16384,0.02629511058330536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,12288,0.0430391099717882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,16384,0.06649777624342176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,65536,0.2381209002600776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,12288,0.054598222176233925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,10240,0.03481600019666884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,12288,0.05156177613470289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,256,0.004939555707905027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,16384,0.055881778399149575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,8192,0.02702933384312524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,10240,0.020277332928445604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,10240,0.04473155405786303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,16384,0.07157422436608209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,7168,0.024497777223587036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,8192,0.015837333268589444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,8192,0.03855999973085191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,12288,0.02144355575243632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,7168,0.014488000008794995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,6144,0.021418665846188862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,7168,0.033928887711630926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,6144,0.013471999930010902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,6144,0.030573334958818223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,5120,0.018358222312397428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,6144,0.030235555436876085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,1536,0.009266667068004608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,4096,0.0155102229780621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,5120,0.02645333276854621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,5120,0.026711111267407734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,4096,0.01109866632355584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,3584,0.01349066694577535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,10240,0.046464890241622925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,3584,0.019882667395803664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,768,0.006663110935025745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,3584,0.010587555666764578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,5120,0.012423111332787408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,3584,0.021037333541446265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,3072,0.010033778018421596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,7168,0.03456000155872769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,3072,0.018852444158660043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,4096,0.022960889670583937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,2560,0.009393778112199571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,2048,0.009295110901196798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,2560,0.017041777571042378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,3072,0.012221333053376941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,8192,0.038823998636669584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,2048,0.013743999931547375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,2048,0.009031111167536842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,2048,0.015507555670208402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,3072,0.01812799937195248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,4096,0.02240000002914005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,1536,0.011426666544543372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,1024,0.0064995553758409284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,1536,0.00868622213602066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,1536,0.013274666335847644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,768,0.005586666779385672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,1024,0.007967111137178209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,1024,0.009473777479595607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,1024,0.011740444435013665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,512,0.004968889057636261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,1536,0.007985778152942657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,512,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,2560,0.010618666807810465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,256,0.004242666479614046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,512,0.010051555931568146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,2560,0.015828443898095023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,256,0.006979555719428592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,128,0.003896000070704354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,768,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,768,0.010401777923107147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,128,0.006931555353932911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,128,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,64,0.0035422220826148987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,128,0.008977777428097194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,32,0.0038924444880750445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,512,0.007320889168315464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,65536,0.1613644493950738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,32,0.006647999915811751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,256,0.00868977771864997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,768,0.008362666600280339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,65536,0.2282631132337782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,65536,0.22930312156677246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,16384,0.04113599989149306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,51200,0.1846319966846042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,51200,0.1808168888092041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,16384,0.059395558304256864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,16384,0.025268443756633337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,16384,0.06424711147944133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,51200,0.12143022484249538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,12288,0.04571466644605001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,12288,0.02109333376089732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,10240,0.025019556283950806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,51200,0.07454311185412936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,10240,0.018981332580248516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,8192,0.0203795548942354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,10240,0.04303644431961907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,12288,0.02999911043379042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,256,0.006991110742092133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,7168,0.017830222845077515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,8192,0.0367253323396047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,12288,0.05037066671583387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,7168,0.029343999094433252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,7168,0.014167111780908374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,6144,0.015925332903862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,10240,0.03912622067663405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,64,0.006664000037643645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,6144,0.012960000170601739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,6144,0.025544888443417017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,6144,0.02930755416552226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,65536,0.09171822335984971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,4096,0.012274666792816587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,5120,0.025848888688617285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,7168,0.03291733397377862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,4096,0.01868711080816057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,3584,0.011075555450386472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,5120,0.014173333843549093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,8192,0.015281778242852954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,3584,0.0169840008020401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,3072,0.009870222045315636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,3584,0.020436444216304354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,5120,0.02201244400607215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,3072,0.010027555955780877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,2560,0.008991999758614434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,3072,0.018294221825069852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,4096,0.022171555293930903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,2560,0.013603555659453074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,2048,0.007714666426181793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,2560,0.016930666234758165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,8192,0.032467554012934365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,3584,0.010398222340477837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,5120,0.011853333148691388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,2048,0.01498222185505761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,3072,0.015457777513398064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,1536,0.010407110883129967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,1536,0.00832088871134652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,4096,0.01072799993885888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,1536,0.013099555340078143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,2560,0.009365333451165093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,768,0.005127999931573868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,1024,0.008625778059164682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,1024,0.011393778026103973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,2048,0.008952889177534316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,1536,0.006816000160243776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,768,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,768,0.010061333576838175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,512,0.007324444750944774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,1024,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,1024,0.007888889147175683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,256,0.006256000035338932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,2048,0.012116444607575735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,768,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,256,0.009009777671760982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,128,0.0038977778620190094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,256,0.0069822221994400024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,512,0.006933333145247565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,128,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,64,0.003855111284388436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,32,0.003887999802827835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,64,0.005988444305128521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,32,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,512,0.00453955555955569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,65536,0.18793155087365043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,128,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,128,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,51200,0.10397155417336358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,51200,0.07365689012739393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,65536,0.12447910838656956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,51200,0.14857155746883816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,51200,0.17351555824279785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,16384,0.024873778223991394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,16384,0.04974933465321859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,12288,0.023565333750512864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,16384,0.061703999837239586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,256,0.004230222354332606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,10240,0.019314666589101154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,65536,0.2203484508726332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,12288,0.038662221696641706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,12288,0.04809777935345968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,8192,0.015974221958054435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,10240,0.033069332440694175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,10240,0.041598220666249595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,8192,0.015040000279744467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,7168,0.014158222410413953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,512,0.009706666900051964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,65536,0.09229244126213922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,12288,0.02033244404527876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,7168,0.025115556187099878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,6144,0.012823111481136747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,7168,0.014146667387750415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,10240,0.018595554762416415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,16384,0.03265777892536587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,6144,0.012795555922720166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,6144,0.0277137772904502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,8192,0.02773688899146186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,5120,0.018907555275493197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,5120,0.02422222163942125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,4096,0.015948444604873657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,4096,0.021075555019908484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,3584,0.00871377769443724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,5120,0.011077333655622272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,6144,0.02199733257293701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,3584,0.010430222584141625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,3584,0.014831110835075378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,5120,0.011757333245542316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,3584,0.019609777463806998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,3072,0.009729777773221334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,3072,0.013560000393125745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,2560,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,3072,0.017665778597195942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,4096,0.010776888993051318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,2048,0.006504888749784893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,2560,0.01607822212908003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,2048,0.010645333263609143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,4096,0.009522666533788046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,1536,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,2048,0.014188443620999655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,2048,0.008840888738632202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,1536,0.008380444513426887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,1024,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,1536,0.01256977766752243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,2560,0.009392889009581672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,1024,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,1024,0.007824000385072496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,768,0.004576000074545543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,1024,0.01075111081202825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,3072,0.008045333127180735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,8192,0.03479289015134176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,768,0.007582222421964009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,768,0.006991110742092133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,768,0.010076444182131026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,256,0.003597333199448056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,512,0.007089777953094906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,1536,0.009381333159075843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,512,0.009706666900051964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,128,0.003549333247873518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,256,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,256,0.008728000024954477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,128,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,64,0.0032097777972618737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,128,0.008319999608728621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,32,0.0031964443624019623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,64,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,32,0.005668444352017508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,65536,0.08938311205969916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,7168,0.03174044357405768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,65536,0.09131644169489543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,128,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,65536,0.21704444620344374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,512,0.003943110919660992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,512,0.006305777778228124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,51200,0.14429244730207655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,16384,0.027753776974148218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,51200,0.07327110899819268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,65536,0.18507199817233613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,16384,0.04840444525082906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,12288,0.02797244323624505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,16384,0.024136000209384497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,12288,0.03796089026663039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,51200,0.17115644613901773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,12288,0.047338667843076915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,10240,0.023750222391552393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,12288,0.01996977792845832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,2560,0.01198755535814497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,8192,0.01964977714750502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,10240,0.018264888061417472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,10240,0.03228977653715346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,10240,0.040553778409957886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,51200,0.07157333029641046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,8192,0.03421511252721151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,7168,0.024243555135197107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,6144,0.015415110521846347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,7168,0.01354311075475481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,7168,0.030557334423065186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,6144,0.012523555921183692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,6144,0.021446221404605444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,5120,0.013531555732091268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,6144,0.02717599935001797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,16384,0.060334225495656334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,4096,0.009203555683294931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,5120,0.018623110320832994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,5120,0.02370133333735996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,4096,0.015858666764365304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,3584,0.008522666990756989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,256,0.005745777653323279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,3584,0.014498665928840637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,3584,0.019476445184813607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,8192,0.014833778142929077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,3072,0.007595555649863348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,5120,0.01146755533085929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,3072,0.009740444521109263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,3072,0.013307555682129331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,2560,0.0069875551594628235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,4096,0.010436444646782346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,3072,0.017197334104114108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,4096,0.020353777541054618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,2560,0.00907911111911138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,2560,0.01184977756606208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,2560,0.01578044394652049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,2048,0.010435555544164447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,8192,0.02693066663212246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,3584,0.010049777726332346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,1536,0.00906399968597624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,1024,0.004559111264016894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,1536,0.012103111379676394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,2048,0.006039999839332368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,1024,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,768,0.004271111140648524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,1024,0.01071999967098236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,2048,0.00868444475862715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,7168,0.017887110511461895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,768,0.007330666813585494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,768,0.009724444813198512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,1536,0.008015111088752747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,512,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,256,0.003560888684458203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,512,0.009242666264375051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,1024,0.007682667010360294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,256,0.005608888963858287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,1536,0.005299555758635203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,128,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,256,0.008346666892369589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,2048,0.014171555638313293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,512,0.0038773334688610504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,128,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,128,0.005621333503060871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,128,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,512,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,64,0.0053839998112784485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,65536,0.14157599873012966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,256,0.006672888994216919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,51200,0.043418665726979576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,64,0.0032417778339650896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,65536,0.215101334783766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,32,0.0032151111712058387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,16384,0.01771288944615258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,51200,0.07311199771033393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,51200,0.11304444736904568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,51200,0.16949333084954155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,65536,0.05191377798716227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,16384,0.05999555852678087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,12288,0.029725333054860432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,768,0.006785777707894643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,10240,0.013841778039932251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,12288,0.04672977659437391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,10240,0.025757332642873127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,10240,0.03910933269394769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,32,0.005614222337802251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,8192,0.013119999733236102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,8192,0.021399999658266704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,65536,0.09073599841859604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,12288,0.019504000743230183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,7168,0.011980444192886353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,7168,0.01943199998802609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,16384,0.038021332687801786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,10240,0.018079110317760043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,6144,0.010645333263609143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,7168,0.01347733371787601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,7168,0.029987557066811457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,8192,0.03329244587156508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,6144,0.01759377784199185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,5120,0.009326222042242685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,6144,0.012455999851226807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,6144,0.026546667019526165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,4096,0.00796977761718962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,5120,0.011419555379284753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,5120,0.015455111861228943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,5120,0.02312533391846551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,16384,0.023450666003757056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,12288,0.015081778168678284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,4096,0.013460444079505073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,3072,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,4096,0.020238222347365487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,3584,0.010048888623714447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,3584,0.012439111040698158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,2560,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,3072,0.011413333316644033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,3072,0.016873778568373788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,8192,0.014481777946154276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,2048,0.005664888769388199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,2560,0.010028444230556488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,2560,0.015511110424995422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,3584,0.007288888924651676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,2048,0.009024000002278222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,2048,0.014124444789356656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,1536,0.008351999852392409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,1536,0.012063110868136087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,1024,0.004330666528807746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,2560,0.00923466682434082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,4096,0.010395555860466426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,1536,0.004945777770545748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,768,0.004230222354332606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,1024,0.007627555893527136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,3072,0.009697777529557547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,768,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,512,0.0035502223504914176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,1536,0.008011555506123437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,512,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,512,0.009032888544930352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,2048,0.0086666668454806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,1024,0.010398222340477837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,3584,0.018983999888102215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,256,0.008009778128729926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,768,0.009727111293209923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,128,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,1024,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,256,0.003546666767862108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,64,0.002895999906791581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,128,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,32,0.0029066666546795103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,128,0.008037333687146505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,256,0.0052915554907586836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,65536,0.040771553913752236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,64,0.004944889081848992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,768,0.0073520003093613525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,51200,0.033088889386918806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,65536,0.09056888686286078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,65536,0.14095377922058105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,65536,0.21469599670834014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,16384,0.013111111190583972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,51200,0.16919555928972033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,16384,0.03738933470514085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,256,0.006676444576846228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,16384,0.022985777921146814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,12288,0.01108977778090371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,12288,0.029218667083316382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,51200,0.07306222120920818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,10240,0.010445333189434476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,12288,0.019476445184813607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,10240,0.025399110383457605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,128,0.0031893334041039148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,16384,0.05946489175160726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,8192,0.009358222285906473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,8192,0.021205332544114854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,512,0.0069857777820693115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,7168,0.008912888665994009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,8192,0.014479110638300577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,7168,0.019345778557989333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,32,0.005246222019195557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,6144,0.00833777752187517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,6144,0.01739199956258138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,8192,0.03297066688537598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,6144,0.012464000119103326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,5120,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,6144,0.02649066680007511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,7168,0.029519110918045044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,7168,0.01347733371787601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,5120,0.0151573336786694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,4096,0.007634667058785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,5120,0.023072888453801472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,51200,0.11045155260297988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,4096,0.013159111142158508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,12288,0.04616444309552511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,3584,0.0069520001610120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,4096,0.02030311193731096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,4096,0.010432889064153036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,3072,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,3584,0.009775111244784461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,3584,0.01236533373594284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,3584,0.018569777409235638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,2560,0.006048888795905643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,10240,0.017986666825082567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,5120,0.011405333048767514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,2560,0.00903911143541336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,2048,0.0052879999081293745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,2560,0.015477332803938123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,2048,0.009110222260157267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,2048,0.013797333670987023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,1536,0.004934222333961063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,3072,0.009456000394291347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,3072,0.016886222693655226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,1536,0.007982222570313348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,1536,0.008003555238246918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,1536,0.012105777859687805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,1024,0.0069919998447100324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,768,0.0036764442920684814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,1024,0.007628444168302748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,10240,0.03913866811328464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,3072,0.011401777466138205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,512,0.003559999995761447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,768,0.006266666783226862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,768,0.009707555174827576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,1024,0.004228444563017951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,512,0.006991110742092133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,512,0.00600355573826366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,512,0.009377777576446533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,256,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,128,0.0032017777363459268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,256,0.00832444429397583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,2048,0.008616888688670265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,128,0.006686222222116258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,128,0.005084444251325396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,64,0.0028497777465316984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,128,0.008021333151393467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,2560,0.010048000348938836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,256,0.003208000005947219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,65536,0.03956088754865859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,64,0.005129777722888523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,51200,0.03427555494838291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,32,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,16384,0.012524444195959302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,12288,0.01072355525361167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,256,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,65536,0.14071999655829534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,51200,0.10998844438129002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,16384,0.03713422351413303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,12288,0.02881422307756212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,768,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,10240,0.02513511147763994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,32,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,8192,0.021310221817758348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,5120,0.007697777615653143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,7168,0.019426667028003268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,4096,0.007638221813572778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,6144,0.017228444417317707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,3584,0.007324444750944774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,5120,0.01534222231970893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,3072,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,4096,0.013220444321632385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,8192,0.009003555609120263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,3584,0.01220533334546619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,2048,0.005250666704442766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,7168,0.008663111262851292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,1536,0.004593777573770947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,6144,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,2560,0.010055999788973067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,1024,0.0038977778620190094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,768,0.003850666599141227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,2048,0.00904799997806549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,512,0.003555555724435382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,1024,0.0069848886794514125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,256,0.0031742221779293488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,768,0.006265777680608962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,1024,0.010728889041476779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,512,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,64,0.002529777793420686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,256,0.005246222019195557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,32,0.002842666581273079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,128,0.0052826665341854095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,65536,0.03579644362131754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,64,0.00498488876554701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,51200,0.029708445072174072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,32,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,16384,0.009374221993817223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,65536,0.1396097739537557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,12288,0.008389333056079017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,51200,0.10963910818099976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,10240,0.010027555955780877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,10240,0.01621866722901662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,16384,0.03700977894994948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,2560,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,8192,0.013812444276279874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,3072,0.01149777736928728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,7168,0.012900444368521372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,6144,0.011734222372372946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,7168,0.019319110446506076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,5120,0.010764444039927589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,128,0.0028702221396896574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,4096,0.009731555150614845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,5120,0.015323556131786771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,3584,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,12288,0.029022223419613306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,10240,0.024865777956114873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,8192,0.02125866711139679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,2560,0.005629333357016246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,1536,0.008006222546100616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,2048,0.005583111196756363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,2560,0.010041777458455827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,1536,0.004559111264016894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,2048,0.00903733323017756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,1024,0.0038764443662431505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,1536,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,768,0.003543111185232798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,1024,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,512,0.0032666667054096856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,256,0.002842666581273079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,3072,0.011391110718250275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,6144,0.01734755602147844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,512,0.005872000008821487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,64,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,256,0.0052675555149714155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,32,0.00254133323000537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,128,0.004970666848950916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,64,0.004933333231343163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,3072,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,3584,0.012425777812798819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,768,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,128,0.0028515555378463534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,4096,0.01333333303531011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,16384,0.5890026622348362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,16384,1.7858577304416234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,16384,1.891816880967882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,12288,1.348515510559082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,12288,0.44043999248080784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,51200,2.1347147623697915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,10240,1.059823989868164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,32,0.004994666824738185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,10240,1.1034684711032443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,8192,0.9694933361477323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,12288,1.427214198642307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,8192,0.8725360234578451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,7168,0.7302426762051053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,51200,5.52564197116428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,7168,0.2829475667741564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,6144,0.5756150881449381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,7168,0.8053075472513834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,5120,0.5712195502387153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,6144,0.6532097922431098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,6144,0.2433662149641249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,5120,0.20444533560011122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,4096,0.4394533369276259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,5120,0.5460249053107368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,4096,0.1720728874206543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,3584,0.4044675562116835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,4096,0.41956978374057347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,3584,0.14897067017025417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,3072,0.30204976929558647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,8192,0.30323555734422475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,3584,0.3892666763729519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,3072,0.13316710789998373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,2560,0.11487644248538548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,2048,0.2365084489186605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,10240,0.3755199909210205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,2560,0.2700951099395752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,1536,0.16893333858913848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,2048,0.0971333318286472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,1536,0.07883644766277738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,1024,0.0619680020544264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,1024,0.13051556216345891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,768,0.10466933250427246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,3072,0.3964373270670573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,512,0.07199733124838935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,768,0.0552151103814443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,2048,0.21925155321756998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,768,0.10558222399817573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,256,0.049806220663918384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,512,0.09096178081300522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,256,0.047305779324637524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,256,0.08234933349821302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,2560,0.2474364439646403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,128,0.04529244369930691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,64,0.03657777773009406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,32,0.040806220637427434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,64,0.07676088809967041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,1024,0.11177955733405219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,32,0.07721600267622206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,1536,0.17313688331180146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,128,0.04230933388074239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,128,0.07669422361585829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,51200,6.737493303087022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,512,0.050550222396850586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,16384,1.258119159274631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,65536,2.1823688083224826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,51200,4.023896111382379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,16384,1.4900764889187281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,16384,0.455227533976237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,51200,1.7155111100938585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,65536,5.148492601182726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,12288,1.010272873772515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,10240,0.877687136332194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,12288,1.0497893227471244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,10240,0.8635404374864367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,8192,0.6366471184624566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,10240,0.2948391172620985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,8192,0.2486462328169081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,12288,0.34612001313103574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,7168,0.2135875489976671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,8192,0.7123635609944662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,65536,6.9389466179741754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,6144,0.4748106532626682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,6144,0.18700710932413736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,5120,0.43139465649922687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,7168,0.5854658020867242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,4096,0.3631546762254503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,6144,0.5018515586853027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,5120,0.4182737668355306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,4096,0.13279022110833064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,7168,0.6133057806226943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,3584,0.3243146737416585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,4096,0.40351555082533097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,3072,0.26352887683444554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,3584,0.2872062259250217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,2560,0.2268622186448839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,3072,0.24899999300638834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,2048,0.18362843990325928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,5120,0.16189689106411403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,2560,0.21536710527208117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,2048,0.08772710959116618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,2048,0.1733271148469713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,1024,0.08162844181060791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,3584,0.12190133995480007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,51200,5.273150973849826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,1024,0.049570666419135205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,768,0.08229066928227742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,1024,0.10326311323377822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,2560,0.09135822455088298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,512,0.05240622162818909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,768,0.045066667927636035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,768,0.08346399996015762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,1536,0.14298310544755724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,256,0.04072711202833388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,512,0.07230933507283528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,256,0.038541333542929754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,1536,0.1412008868323432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,64,0.03285866644647386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,128,0.06128355529573229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,3072,0.11053688658608331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,32,0.03398399882846408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,64,0.06131111250983345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,32,0.061666667461395264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,512,0.04153955645031399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,128,0.034254223108291626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,1536,0.06404444244172838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,128,0.03700177868207296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,256,0.06563999917772081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,65536,0.657425774468316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,65536,1.743262184990777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,16384,0.38739644156561953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,16384,0.4415866798824734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,51200,0.522997326321072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,12288,0.32339999410841197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,12288,0.12622843848334417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,10240,0.3107137680053711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,65536,2.037687089708116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,12288,0.3297084437476264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,51200,1.6621501710679796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,8192,0.25270401106940377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,10240,0.2762320041656494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,8192,0.08552533388137817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,8192,0.22461512353685167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,51200,1.3491866853502061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,7168,0.07431644201278687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,16384,0.16354133023156062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,7168,0.20704800552792021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,5120,0.12829955418904623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,6144,0.06566311253441705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,4096,0.10558933019638062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,5120,0.05612622367011177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,5120,0.13825599352518717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,3584,0.10778044329749213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,4096,0.04693866769472758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,10240,0.10688800281948513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,4096,0.11216710673438178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,3072,0.08737066719267104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,2560,0.07798755831188627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,3072,0.08594577842288548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,2560,0.07418400049209595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,2048,0.05964710977342394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,3584,0.042111999458736844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,7168,0.1944826708899604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,2048,0.02881866693496704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,1536,0.04792444573508369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,2560,0.03315911028120253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,6144,0.16467377874586317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,1536,0.049044443501366504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,1024,0.03675999906327989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,1024,0.020046222541067336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,1024,0.0373599992858039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,3584,0.09849244356155396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,6144,0.1621626615524292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,768,0.018392889036072623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,512,0.021586666504542034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,512,0.016540444559521146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,256,0.015784000356992085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,2048,0.061612447102864586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,256,0.022674666510687932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,768,0.02832888894610935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,3072,0.037610666619406805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,256,0.015203555425008139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,64,0.011048888994587792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,768,0.030287110143237646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,128,0.02204533252451155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,32,0.011401777466138205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,64,0.021692444880803425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,32,0.02166755497455597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,128,0.01310755560795466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,128,0.015255111787054272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,65536,0.4851359791225857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,65536,1.295325385199653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,512,0.02546577817863888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,51200,0.3867066701253255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,65536,1.65338929494222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,16384,0.37703821394178605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,12288,0.24624178144666883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,51200,1.1595448387993705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,10240,0.21043645011054146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,12288,0.2536373403337267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,51200,0.9034356011284722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,10240,0.08392977714538574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,8192,0.14140978124406603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,7168,0.12417244911193848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,16384,0.11965778138902451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,8192,0.16506489117940268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,7168,0.05673955546485054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,6144,0.10563466946283977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,12288,0.0928000013033549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,5120,0.11000621981090969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,6144,0.050147556596332125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,6144,0.12614933649698892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,1536,0.023751111494170293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,16384,0.36155109935336643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,4096,0.091011557314131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,8192,0.0645244452688429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,5120,0.10816622442669338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,4096,0.08629600207010905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,3584,0.03274755676587423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,3072,0.07196355528301664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,3584,0.07658221986558703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,7168,0.1463599999745687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,2560,0.05891911188761393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,3072,0.02923733327123854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,2048,0.04928266670968798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,3072,0.06720266739527385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,2560,0.05782044596142239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,10240,0.20925511254204643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,5120,0.04291199975543552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,2048,0.04838844471507602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,4096,0.03598577777544657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,1536,0.019694222344292533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,1536,0.03870666689342923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,768,0.019080888893869188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,1024,0.016718222035302054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,1024,0.029767112599478826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,512,0.015079999963442484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,768,0.01499555508295695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,768,0.024159999357329473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,2560,0.026775999201668635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,256,0.011748444702890186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,512,0.01387644476360745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,512,0.02035466664367252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,1536,0.03297066688537598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,256,0.012916444076432122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,3584,0.07980710930294461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,256,0.01828799976242913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,1024,0.02516266703605652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,128,0.013167111410035027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,128,0.017786666750907898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,64,0.01756622228357527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,32,0.017551110850440133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,2048,0.023025777604844835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,65536,0.9864675733778211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,64,0.00906577789121204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,32,0.009124444590674506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,16384,0.24653866555955675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,65536,1.2858986324734158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,51200,0.3473466767205133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,12288,0.1954195499420166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,16384,0.10606577661302354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,51200,1.0097039540608723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,12288,0.08367644415961371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,65536,0.4313066535525852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,12288,0.21429422166612414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,51200,0.831241766611735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,10240,0.06867555777231853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,10240,0.1845137808057997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,7168,0.13596799638536242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,8192,0.055719110700819224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,8192,0.14123466279771593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,16384,0.2851751115587023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,6144,0.11431911256578232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,7168,0.04954933457904392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,10240,0.17363999949561226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,5120,0.10319733619689941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,6144,0.10819466908772786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,5120,0.03781866696145799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,4096,0.08242933617697822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,8192,0.1598328881793552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,3584,0.07271111011505127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,4096,0.0742924412091573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,3584,0.06511733267042372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,128,0.010064889159467485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,6144,0.044439110491010875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,3072,0.02602844436963399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,2560,0.052729778819613986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,5120,0.09149510992897882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,4096,0.031566222508748375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,2560,0.02347022294998169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,2560,0.049469331900278725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,2048,0.02073777715365092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,2048,0.041674666934543185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,7168,0.1254968908098009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,1536,0.017533333765135873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,1536,0.03346577617857192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,1024,0.01495822270711263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,1024,0.025974222355418738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,2048,0.041096889310412936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,768,0.013815110756291283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,768,0.021568889419237774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,1536,0.029353777567545574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,3584,0.029342220889197454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,256,0.010375111467308467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,3072,0.05948978000217014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,1024,0.02160355614291297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,512,0.018175111876593698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,128,0.009357333183288574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,256,0.012006222373909421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,64,0.008341333104504479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,3072,0.057357331116994224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,768,0.017535999417304993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,32,0.008416888614495596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,64,0.015527110960748462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,32,0.015503999259736804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,512,0.01275199982855055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,65536,0.8186204698350695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,256,0.016015110744370353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,128,0.012086222569147745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,128,0.0158595558669832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,51200,0.6831750869750977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,65536,1.0732275644938152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,512,0.014020444618331062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,16384,0.23086576991611055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,12288,0.18051200442843965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,51200,0.7966924243503146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,10240,0.15182667308383518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,12288,0.17137955294715035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,65536,0.3609022299448649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,8192,0.1156960063510471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,10240,0.05932088692982992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,10240,0.14313688543107775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,7168,0.11598400274912517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,8192,0.04649244414435493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,51200,0.2775431209140354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,7168,0.04247555467817518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,7168,0.10269066360261704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,6144,0.037069333924187556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,5120,0.07205600208706327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,6144,0.08944444523917304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,12288,0.06794399685329862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,4096,0.067793779902988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,5120,0.07536977529525757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,3584,0.06172177526685926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,4096,0.06189066833919949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,8192,0.11700799730088975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,3584,0.025014221668243408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,3584,0.054535110791524254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,16384,0.08996799919340347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,2560,0.04569155640072293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,3072,0.048078222407235034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,4096,0.027968890137142603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,16384,0.2369635634952121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,2560,0.020802666743596394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,2560,0.041724443435668945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,3072,0.05603910817040337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,2048,0.018703111343913607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,2048,0.035240888595581055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,1024,0.019927110936906602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,1536,0.028720001379648846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,3072,0.023039999935362074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,5120,0.03259199857711792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,768,0.015579554769727917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,1024,0.022152000003390845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,768,0.01828888886504703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,768,0.012435555458068848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,512,0.011932444241311816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,1536,0.028547555208206177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,256,0.009124444590674506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,512,0.011705777711338468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,512,0.015507555670208402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,1536,0.015970667203267414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,128,0.007648888561460707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,256,0.013952889376216464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,64,0.006513777706358168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,128,0.013454222016864352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,32,0.006618666566080517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,64,0.013225778109497495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,32,0.013467555244763693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,256,0.011071999867757162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,1024,0.013576000101036496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,65536,0.8267688751220703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,65536,0.3214755588107639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,51200,0.5842320124308268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,65536,0.927853372361925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,51200,0.24642133712768555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,12288,0.17410043875376383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,16384,0.081094225247701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,51200,0.6917440096537272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,16384,0.2079484462738037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,12288,0.061568001906077065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,128,0.010983111129866706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,12288,0.15228354930877686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,8192,0.1039280030462477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,10240,0.12807288434770372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,2048,0.03687999977005853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,8192,0.04127111037572225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,7168,0.0828862190246582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,8192,0.10369155804316203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,6144,0.07335288657082452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,7168,0.03728444377581278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,5120,0.06743288702434964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,6144,0.03349422083960639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,6144,0.079276442527771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,10240,0.12033689022064209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,5120,0.06753689050674438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,10240,0.05140622125731575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,3584,0.04711822337574429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,6144,0.08834311034944321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,16384,0.22331555684407553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,3584,0.022942221826977197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,3072,0.042938666211234204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,3584,0.04874044325616625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,7168,0.09154133664237128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,2560,0.034128000338872276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,3072,0.020770667327774894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,2048,0.028900444507598877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,3072,0.04320266511705187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,4096,0.05506755577193367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,1536,0.022574222750133936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,2048,0.016805332567956712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,2048,0.03169511093033685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,4096,0.024832000335057575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,1536,0.014560000763999091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,1536,0.025839999318122864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,768,0.014236445228258768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,1024,0.01981511049800449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,768,0.011752888560295105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,512,0.011515555282433828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,5120,0.029150221082899306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,2560,0.03769422239727444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,512,0.01093777765830358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,512,0.014142221874660917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,4096,0.055009779002931386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,128,0.007337777978844113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,256,0.010458666417333813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,1024,0.012705778082211813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,64,0.006267555471923616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,128,0.010397333237859938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,128,0.012439111040698158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,768,0.01647644407219357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,2560,0.019134221805466544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,64,0.012427555190192329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,32,0.01240355521440506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,256,0.009031111167536842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,1024,0.017276444368892245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,65536,0.5991911358303493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,65536,0.2590542104509142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,256,0.012825777961148156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,16384,0.17097066508399117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,51200,0.2095573345820109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,12288,0.11306756072574192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,16384,0.0651600029733446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,51200,0.5895937813652886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,10240,0.10515199767218696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,12288,0.052765332990222506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,12288,0.13288977411058214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,8192,0.08884355756971572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,10240,0.04464088877042135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,10240,0.11342310905456543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,8192,0.03569599986076355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,8192,0.09141600131988525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,51200,0.4287724494934082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,7168,0.03254666593339708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,7168,0.08107644319534302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,65536,0.7806017663743762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,6144,0.029327998558680218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,6144,0.06999822457631429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,5120,0.02515911062558492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,4096,0.04626133375697666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,7168,0.07586222224765353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,4096,0.021912889348136053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,6144,0.06508266925811768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,4096,0.04845866560935974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,32,0.00666133314371109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,3072,0.03350044290224711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,3584,0.043339557117886014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,3072,0.018681777848137748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,2560,0.028195556667115953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,5120,0.05997688902748955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,2560,0.01685422162214915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,2048,0.024015110399987962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,3584,0.01976355579164293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,16384,0.17839645014868843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,2048,0.029104888439178467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,1536,0.019298666053348117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,2048,0.014928888943460254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,1024,0.01496266656451755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,1536,0.013213333156373767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,3072,0.03836266530884637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,1536,0.023423999547958374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,768,0.012143111063374413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,1024,0.011477332976129321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,1024,0.017702221870422363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,2560,0.0332897769080268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,768,0.01517599986659156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,256,0.007673777639865875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,512,0.012902222573757172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,128,0.0063164445261160536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,256,0.009405333134863112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,256,0.011756444142924415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,64,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,128,0.011471110913488599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,32,0.005691555639108022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,512,0.00941066692272822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,768,0.010755555497275459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,3584,0.03934577769703335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,512,0.010050666828950247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,65536,0.5626364284091526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,65536,0.2411359945933024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,128,0.009728000395827824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,65536,0.6452364391750759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,16384,0.14632266097598606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,51200,0.19035466512044272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,64,0.011394666300879585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,16384,0.06161955330106947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,51200,0.5422382354736328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,16384,0.15685866938696966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,10240,0.09298577573564316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,12288,0.11543288495805527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,5120,0.056419557995266385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,8192,0.07309599717458089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,10240,0.09688799911075169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,51200,0.42906665802001953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,7168,0.06326399909125434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,8192,0.07897066407733493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,7168,0.030823998981051978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,7168,0.06951022148132324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,12288,0.1083022223578559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,32,0.01128088848458396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,5120,0.04640444450908237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,6144,0.027499554885758296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,12288,0.048476444350348584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,4096,0.039255112409591675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,5120,0.024559999505678814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,10240,0.043615112702051796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,4096,0.02088800072669983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,4096,0.04275022281540764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,8192,0.03377155462900797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,3584,0.019355555375417072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,3584,0.03851111067665948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,6144,0.05611377954483032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,3072,0.018012444178263348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,3072,0.03400977783732944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,2048,0.020232889387342665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,2560,0.029642668035295274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,2048,0.01460533340771993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,5120,0.05186133252249824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,1536,0.016743110285864938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,3072,0.029674665795432195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,1024,0.012162666353914471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,1536,0.020997333857748244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,2560,0.025272000167104933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,768,0.01015644437736935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,1024,0.011392888923486074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,2560,0.01605244477589925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,512,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,768,0.010504888991514841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,768,0.013607111242082385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,256,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,512,0.009732444253232744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,512,0.011762667033407422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,256,0.009397333694828881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,1536,0.012828444441159567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,128,0.010384889112578498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,1024,0.0158595558669832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,64,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,6144,0.06160800324545967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,32,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,2048,0.025257777836587694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,64,0.010079999764760336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,128,0.006229333165619109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,32,0.010060444474220276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,256,0.010483555495738983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,65536,0.36699289745754665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,51200,0.3221653302510579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,16384,0.10115200281143188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,51200,0.1571520037121243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,128,0.009348444640636444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,51200,0.4252053366767035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,12288,0.08381688594818115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,3584,0.03357777661747403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,10240,0.07595377498202853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,12288,0.039677331844965615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,12288,0.09740622176064385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,10240,0.03445333242416382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,10240,0.08232088883717854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,65536,0.20068800449371338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,7168,0.049238221512900464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,8192,0.02807733416557312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,65536,0.5457439952426487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,6144,0.0477422210905287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,7168,0.0258568889564938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,5120,0.04019999835226271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,16384,0.05017866690953573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,16384,0.12753955523173013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,6144,0.023399111297395494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,4096,0.032046222024493747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,5120,0.04447288976775276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,3584,0.028901331954532202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,4096,0.018072000808186002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,4096,0.03705688979890611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,3584,0.016857778032620747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,8192,0.06708711385726929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,3584,0.03347999850908915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,2560,0.022791999909612868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,7168,0.05944266584184435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,3072,0.02988977895842658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,2048,0.017861333158281114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,2560,0.014092443717850579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,5120,0.02076088885466258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,2560,0.02590488890806834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,1536,0.014763555592960782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,2048,0.012930666406949362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,8192,0.06438044706980388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,2048,0.02221777704026964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,1536,0.011425777441925473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,768,0.009062222308582729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,1536,0.018279999494552612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,1024,0.010055111514197456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,512,0.007273777491516537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,768,0.009411555197503831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,768,0.012169777519173093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,3072,0.0155102229780621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,512,0.010680889089902243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,6144,0.051826665798823036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,256,0.009664889011118147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,1024,0.010760000182522668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,128,0.008592888712882996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,3072,0.024695111645592585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,1024,0.014096000128322177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,32,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,256,0.0061048890153567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,32,0.00902133352226681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,256,0.008634666601816813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,128,0.004954666727119022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,65536,0.18483289082845053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,51200,0.35184354252285427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,65536,0.4802257749769423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,16384,0.10843822028901841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,51200,0.15247821807861328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,51200,0.36650933159722227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,16384,0.050112889872656934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,12288,0.0835004448890686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,65536,0.3764799965752496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,10240,0.06649600134955512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,12288,0.08732799688975017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,64,0.00489688871635331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,8192,0.05483999848365784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,10240,0.03410222133000692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,10240,0.07251377900441487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,512,0.009012444151772393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,8192,0.059188445409139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,16384,0.11549866199493408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,7168,0.025592888394991558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,7168,0.0528400010532803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,12288,0.03923377725813124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,5120,0.034493333763546415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,128,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,6144,0.04594044552909004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,4096,0.02868088748719957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,5120,0.020678222179412842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,5120,0.03970311085383097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,7168,0.05046844482421875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,4096,0.017487999465730455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,4096,0.03279377685652839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,3584,0.015951111912727356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,6144,0.04018488857481215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,3584,0.029672000143263075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,6144,0.022871999277008906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,2560,0.018950222267044913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,3072,0.026304889056417678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,2560,0.013625777430004544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,64,0.009212444225947062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,2048,0.015983111328548856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,3584,0.025685333543353613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,1536,0.01294133315483729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,2048,0.012667555775907306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,2048,0.019260444574885897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,1024,0.009687999884287516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,1536,0.016372443901167977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,3072,0.015106666419241162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,768,0.008574222524960836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,1024,0.012719111310111152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,768,0.009338666995366415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,8192,0.02795377704832289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,512,0.007300444775157505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,512,0.008911111288600499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,256,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,3072,0.021988444858127173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,128,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,256,0.008384000096056197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,256,0.008972444468074376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,64,0.004905777672926585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,128,0.008381333616044786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,32,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,768,0.01109155515829722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,64,0.00869066682126787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,32,0.008667555948098501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,512,0.009752000371615091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,65536,0.35872001118130153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,1536,0.011247999966144562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,65536,0.15767733256022134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,1024,0.009995555712117089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,16384,0.09186666541629368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,51200,0.3406906657748752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,16384,0.040800889333089195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,12288,0.06915199756622314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,2560,0.023131555981106226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,16384,0.10089688830905491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,10240,0.05468711256980896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,12288,0.032391998502943255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,12288,0.07624533441331652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,10240,0.031186666753556993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,10240,0.06525599956512451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,7168,0.04200088977813721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,51200,0.2902444468604194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,8192,0.052912887599733144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,6144,0.035791999763912626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,7168,0.04744355546103584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,6144,0.018852444158660043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,6144,0.04151022103097703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,5120,0.030174222257402208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,8192,0.04713244570626152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,4096,0.023399111297395494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,5120,0.03569422165552775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,8192,0.02304711110062069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,3584,0.020706666840447318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,65536,0.4164942105611165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,4096,0.014526221487257214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,4096,0.029959999852710303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,3072,0.019080888893869188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,3584,0.014093332820468478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,128,0.00869333330127928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,3072,0.013223110801643796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,3072,0.023822221491071913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,2048,0.013598221871587964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,2560,0.02068355513943566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,5120,0.016864889197879367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,2048,0.010585777461528778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,2048,0.01777066621515486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,51200,0.12344000074598525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,1536,0.009719111025333405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,1024,0.008661333057615492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,7168,0.02370311154259576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,1536,0.014831110835075378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,2560,0.016213332613309223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,3584,0.02723111046685113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,1024,0.011750222080283694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,768,0.00868266655339135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,2560,0.011414222419261932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,256,0.005345777918895085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,512,0.008117333054542542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,512,0.009249777429633671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,128,0.004563555535342959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,256,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,256,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,64,0.00444088876247406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,128,0.007637333538797166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,32,0.004231111043029361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,128,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,768,0.007601777712504069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,1024,0.009038222332795462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,64,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,512,0.006330666856633292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,65536,0.3044702212015788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,768,0.010281778044170803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,51200,0.2043377823299832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,65536,0.14595733748541936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,65536,0.3594542344411214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,51200,0.1126577721701728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,12288,0.05766310956743029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,16384,0.037816888756222196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,16384,0.08903021944893731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,12288,0.03020711077584161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,10240,0.05127200153138903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,12288,0.06815466615888807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,8192,0.037773334317737155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,32,0.008157333566082848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,10240,0.057589332262674965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,7168,0.035762667655944824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,8192,0.022210665875011023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,16384,0.07301777601242065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,1536,0.011063111325105032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,6144,0.02693066663212246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,7168,0.0203840004073249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,51200,0.2848871019151476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,7168,0.04230133361286587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,6144,0.018558222386572096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,4096,0.01962933275434706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,6144,0.037224001354641385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,5120,0.03222666515244378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,3584,0.0181057784292433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,10240,0.02721333338154687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,4096,0.02713511056370205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,3072,0.016428444120619033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,3584,0.024275556206703186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,8192,0.047672000196244985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,2560,0.014427555931939019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,3072,0.01219555570019616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,5120,0.024865777956114873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,2560,0.018921777606010437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,5120,0.01628888812330034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,1536,0.010268444816271463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,2048,0.010392000277837118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,2048,0.016199999385409884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,1024,0.007982222570313348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,1536,0.009681777821646797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,3584,0.013106666505336761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,768,0.006950221955776215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,1024,0.010779555473062726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,3072,0.02163733376397027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,768,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,2048,0.012459555433856117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,512,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,256,0.005221333354711533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,4096,0.014367999302016364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,256,0.007677333222495184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,1536,0.013614222407341003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,256,0.007727111379305522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,1024,0.008984000318580205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,64,0.003909333298603694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,128,0.007634667058785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,32,0.0042453333735466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,2560,0.011418666276666852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,512,0.005971555494599872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,64,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,65536,0.24856532944573295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,32,0.007667555577225155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,65536,0.13702221711476645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,65536,0.31183910369873047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,128,0.004563555535342959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,51200,0.10850755373636882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,51200,0.24801688724093965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,12288,0.04766311248143514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,16384,0.03649155630005731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,16384,0.07761955261230469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,10240,0.03966933488845825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,12288,0.029581334855821397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,12288,0.059843553437127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,512,0.008489777644475302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,10240,0.02659999993112352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,51200,0.16473421785566542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,7168,0.02544266647762722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,8192,0.02160533269246419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,16384,0.06100444661246406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,6144,0.02315644423166911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,7168,0.0195795562532213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,7168,0.037689778539869524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,128,0.007800000409285228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,768,0.009619555539555019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,6144,0.033315555916892156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,4096,0.01644799941115909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,8192,0.03198044498761495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,5120,0.015811555915408664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,10240,0.05117422342300415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,3584,0.015264888604482016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,4096,0.01369599997997284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,3072,0.013711999687883588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,3584,0.012813333008024426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,3584,0.021609778205553692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,2560,0.011743110915025076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,5120,0.01982399986849891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,3072,0.019278221660190158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,2560,0.011071999867757162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,2048,0.009961778091059791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,5120,0.028738667567571003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,4096,0.024109333753585815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,1536,0.008355555435021719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,2048,0.014821334017647637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,1536,0.00943377779589759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,3072,0.012072888513406118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,1024,0.006884444504976273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,1536,0.012636444634861417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,6144,0.01774311065673828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,768,0.006224888894293044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,1024,0.008780444661776224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,2560,0.016953777935769822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,768,0.00869422240389718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,512,0.008006222546100616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,2048,0.010088889135254754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,256,0.0042453333735466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,256,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,128,0.0042160000238153665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,512,0.005127111242877113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,128,0.007000000112586551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,64,0.003682666768630346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,128,0.00758755538198683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,32,0.003648888733651903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,512,0.008017777568764156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,8192,0.04201066825124952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,64,0.007252444823582967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,65536,0.21716977490319145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,32,0.00703288863102595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,65536,0.09305777814653184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,256,0.007360888852013483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,16384,0.056665778160095215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,65536,0.2641448974609375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,12288,0.0412986675898234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,16384,0.025105777713987563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,16384,0.06641511122385661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,10240,0.035657776726616755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,12288,0.05108888943990072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,10240,0.043318221966425575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,8192,0.02725511127048069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,51200,0.07344800233840942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,51200,0.212717334429423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,8192,0.03626399901178148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,8192,0.01591822173860338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,12288,0.021331555313534204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,7168,0.016184889607959323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,7168,0.032720890310075544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,6144,0.013273778061072031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,6144,0.02882044514020284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,51200,0.17330311404334175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,5120,0.012159110771285163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,4096,0.015831111205948722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,7168,0.023990222149425085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,4096,0.011065777805116443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,6144,0.021682666407691106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,3584,0.014275555809338888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,3584,0.011126221881972419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,10240,0.02165422174665663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,3072,0.012792000340090858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,5120,0.01869511107603709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,2560,0.011128000087208219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,3072,0.010737777584128909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,3072,0.016938666502634685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,1024,0.010047111246320937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,2560,0.009664889011118147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,2560,0.014869333969222175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,1536,0.008016889293988546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,2048,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,4096,0.020640888147883944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,1024,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,1536,0.00866844422287411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,1536,0.011159111228254108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,768,0.005961777849329843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,3584,0.018935999936527677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,1024,0.00904444439543618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,512,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,768,0.008047999607192146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,768,0.008347555167145198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,256,0.004563555535342959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,512,0.007359111474619971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,512,0.007345777418878343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,128,0.004224888980388641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,256,0.006991110742092133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,256,0.006816000160243776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,64,0.0038693332009845306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,1024,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,32,0.0036559998989105225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,768,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,5120,0.024481778343518574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,32,0.006587555425034628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,2048,0.01310222182008955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,51200,0.11838666598002116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,128,0.006994666324721442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,128,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,16384,0.035643554396099515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,64,0.006516444600290722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,2048,0.009572444690598382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,16384,0.024148444334665935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,12288,0.026597332623269822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,16384,0.05516977773772346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,65536,0.08609600199593438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,12288,0.02015822297996945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,10240,0.022241777843899194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,65536,0.21146044466230604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,8192,0.018382221460342407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,10240,0.0181368887424469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,10240,0.03644444545110067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,51200,0.06997333632575141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,8192,0.014711999230914645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,51200,0.16815288861592612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,8192,0.030351112286249798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,7168,0.013737777868906656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,7168,0.02718933257791731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,5120,0.012860444684823355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,6144,0.023976888921525743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,5120,0.011736888852384357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,5120,0.02072888943884108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,7168,0.016470222009552848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,4096,0.01758399936887953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,6144,0.01479111115137736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,3584,0.010052444206343757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,4096,0.010434666441546546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,3072,0.009007111191749573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,3584,0.010052444206343757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,3584,0.01607822212908003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,2560,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,3072,0.014601776997248331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,4096,0.010593777729405297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,2048,0.006984000404675801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,2560,0.012816888590653738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,1536,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,2048,0.008681777450773451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,2048,0.011438222395049201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,65536,0.1333084503809611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,1536,0.009691555466916826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,6144,0.012753778033786349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,768,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,1024,0.007639110916190677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,1024,0.008047999607192146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,512,0.0042275554604000514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,12288,0.04243110948138767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,768,0.007486222518814935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,256,0.003576888806290097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,768,0.007328000333574083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,512,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,128,0.003579555700222651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,512,0.006635555376609166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,256,0.006959999601046245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,64,0.003228444399105178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,256,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,128,0.0068408888247278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,32,0.0031884445084465873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,128,0.005982222242487802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,65536,0.11592266294691299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,64,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,32,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,3072,0.009729777773221334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,65536,0.08584622542063396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,65536,0.1729146639506022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,51200,0.06951466533872816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,2560,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,12288,0.024526221884621516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,16384,0.023636443747414485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,16384,0.04532444477081299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,10240,0.02164266672399309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,1024,0.005127999931573868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,12288,0.03569422165552775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,8192,0.018246221873495314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,10240,0.030579557021458943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,10240,0.018246221873495314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,7168,0.016452444924248587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,16384,0.03299377693070306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,8192,0.025774222281244066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,6144,0.014497778481907315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,7168,0.013654222091039022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,51200,0.13760444853040907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,5120,0.01240355521440506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,6144,0.012714666624863943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,6144,0.02032800018787384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,1536,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,5120,0.01167022188504537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,7168,0.02312533391846551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,5120,0.017954667409261067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,3584,0.009882666998439366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,4096,0.010417777631017896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,3072,0.009010666774378883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,3584,0.010085333552625444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,8192,0.01479288935661316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,2560,0.007979555262459649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,3072,0.012563555604881711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,2560,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,2560,0.011406222151385413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,2048,0.007344000041484833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,12288,0.019734222027990553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,1536,0.006310222049554189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,2048,0.008831111093362173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,2048,0.010265777508417765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,1024,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,1536,0.008950221869680617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,3584,0.013792888985739814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,3072,0.009732444253232744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,768,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,1024,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,1024,0.007348444726732042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,512,0.0042275554604000514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,768,0.007373332977294922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,768,0.006666666517655055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,256,0.0035626664757728577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,512,0.006311111152172089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,1536,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,256,0.0069520001610120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,128,0.003539555602603488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,128,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,64,0.0032177778581778207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,4096,0.010567111273606619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,64,0.005624889085690181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,32,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,65536,0.08709422085020278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,4096,0.015152000718646579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,32,0.005607999861240387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,51200,0.07184355788760714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,65536,0.08569688929451837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,65536,0.16575911309983996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,16384,0.02737599942419264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,51200,0.13178577688005236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,16384,0.023022222850057814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,512,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,12288,0.01930488811598884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,51200,0.08927644623650445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,10240,0.01956355571746826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,10240,0.017659554878870647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,8192,0.016598222984208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,10240,0.029734220769670274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,51200,0.06923733154932658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,12288,0.02123111155298021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,16384,0.04391111267937554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,7168,0.013296888934241401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,7168,0.02216266592343648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,128,0.005647111270162795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,6144,0.012413333687517377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,5120,0.011713777979214987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,8192,0.024674667252434626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,5120,0.01145600030819575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,5120,0.01737155516942342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,6144,0.013622221847375235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,4096,0.010103111465771994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,3584,0.008609777523411645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,7168,0.015767999821239047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,3584,0.009691555466916826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,3072,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,4096,0.010010666317409938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,256,0.00582755563987626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,2560,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,3072,0.009331555830107795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,3072,0.012431110772821637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,4096,0.014884443746672736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,2560,0.009001777403884465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,2560,0.011123555401961008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,1536,0.005583999885453119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,2048,0.008452444440788692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,8192,0.014504000544548035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,1024,0.004582222137186262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,1536,0.008056888977686564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,1536,0.008722667064931657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,768,0.00425866660144594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,1024,0.007646222081449296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,12288,0.034620443979899086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,1024,0.0074311114019817775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,512,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,768,0.007292444507280986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,256,0.003524444583389494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,512,0.006979555719428592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,512,0.005991111199061076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,128,0.003248888792263137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,256,0.006611555400821898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,256,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,3584,0.013804444836245643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,64,0.003363555504216088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,32,0.0028773333049482773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,128,0.005611555443869696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,65536,0.04775733417934842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,64,0.005392888767851724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,32,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,51200,0.04233955674701267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,768,0.006676444576846228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,65536,0.12748000356886122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,16384,0.015837333268589444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,51200,0.06884266932805379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,51200,0.09825244214799668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,2048,0.010098666780524785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,128,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,16384,0.02241955531968011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,10240,0.013519110778967539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,12288,0.019000000423855253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,12288,0.026926222774717543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,8192,0.01315111087428199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,10240,0.023432888918452795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,65536,0.08542399936252171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,2048,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,7168,0.01184266640080346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,8192,0.014215999179416232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,7168,0.013227555486891003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,12288,0.013775111072593264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,7168,0.018000000052981906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,6144,0.012278222375445895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,5120,0.009341333475377824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,10240,0.017613333132531907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,5120,0.011233777635627322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,4096,0.008011555506123437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,8192,0.01965688831276364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,4096,0.010064889159467485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,6144,0.010665777656767102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,16384,0.03401511245303684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,3584,0.009736888938479954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,3072,0.006648889018429651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,6144,0.016189333465364244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,3584,0.01165244479974111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,2560,0.006204444501135085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,3072,0.009376000199053023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,2048,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,2560,0.009016888837019602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,5120,0.014322666658295525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,2560,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,1536,0.004910222358173794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,2048,0.008426666259765625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,1024,0.004214222232500712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,2048,0.0086666668454806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,1536,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,3584,0.007674666742483775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,768,0.0038657776183552216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,1536,0.007784888976150089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,1024,0.007616000043021307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,512,0.003585777762863371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,1024,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,256,0.003573333223660787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,768,0.006026666611433029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,512,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,256,0.006653333289755716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,6144,0.019706666469573975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,256,0.005304000029961268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,64,0.0028657778683635923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,128,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,32,0.0029022221763928733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,4096,0.012443555725945367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,64,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,65536,0.03532088796297709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,32,0.004971555537647671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,768,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,51200,0.028974221812354192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,512,0.0069520001610120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,16384,0.013148444394270578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,65536,0.12433066633012559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,51200,0.06848444541295369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,3072,0.010436444646782346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,51200,0.09819644689559937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,16384,0.02209866709179348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,10240,0.01054222219520145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,12288,0.018935999936527677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,12288,0.026087111896938745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,8192,0.010033778018421596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,10240,0.017627555463049147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,10240,0.022875555687480505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,7168,0.00905777762333552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,8192,0.014146667387750415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,65536,0.08482578065660264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,7168,0.013136000268989138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,128,0.002879111096262932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,12288,0.011063111325105032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,6144,0.012087110843923358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,6144,0.016214221715927124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,128,0.0053093334039052325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,5120,0.011071999867757162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,4096,0.007680888805124495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,5120,0.014225777652528552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,3584,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,4096,0.012416000167528788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,8192,0.019306666321224637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,6144,0.009375111096435124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,3584,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,3584,0.011491555306646558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,2560,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,3072,0.009363555245929295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,16384,0.03364533185958862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,3072,0.01052177780204349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,2048,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,2560,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,2560,0.009380444056457942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,2048,0.008671110702885522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,2048,0.008676444490750631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,4096,0.010057777994208867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,1024,0.0042364444169733255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,1536,0.007661333514584436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,768,0.00388355553150177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,1024,0.007654222349325816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,7168,0.017847110827763874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,512,0.003370666669474708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,5120,0.009092444347010719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,512,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,1536,0.004584889031118817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,256,0.0032666667054096856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,1536,0.008021333151393467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,256,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,256,0.006676444576846228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,128,0.003168888803985384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,1024,0.006660444455014334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,64,0.0028977776981062363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,768,0.007273777491516537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,128,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,32,0.002833777831660377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,128,0.00498488876554701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,65536,0.03314933180809021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,51200,0.027523555689387854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,64,0.004962666581074397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,512,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,65536,0.12425422668457031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,12288,0.010416000253624385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,16384,0.03341866532961527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,10240,0.009434666898515489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,12288,0.026297777891159058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,768,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,10240,0.02272355556488037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,7168,0.008103999826643202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,32,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,8192,0.019278221660190158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,7168,0.017768000562985737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,51200,0.09746933645672268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,4096,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,3072,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,3584,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,5120,0.014149333039919535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,3072,0.006401777681377199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,3584,0.011369778050316704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,2560,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,3072,0.010434666441546546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,6144,0.008998221821255153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,2048,0.0052675555149714155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,1536,0.004562666846646203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,5120,0.00833422193924586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,1024,0.0038773334688610504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,1536,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,768,0.003564444267087512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,8192,0.00868000007337994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,1024,0.006635555376609166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,16384,0.012121777567598553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,256,0.003185777821474605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,512,0.005653333332803514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,128,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,2560,0.009366222553782994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,64,0.0028515555378463534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,6144,0.015855999456511605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,32,0.002529777793420686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,128,0.004979555391603046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,65536,0.03198933270242479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,4096,0.012264000044928657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,512,0.0035173334181308746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,32,0.00490399988161193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,16384,0.009703111317422655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,768,0.005998222364319696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,65536,0.12398755550384521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,51200,0.09783466657002766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,10240,0.010778666370444827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,256,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,2048,0.008630221916569604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,8192,0.010088889135254754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,7168,0.009700444837411245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,10240,0.023039110832744177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,64,0.004943111290534337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,6144,0.010025777750545079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,8192,0.019490665859646268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,5120,0.009325332939624786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,7168,0.017829333742459614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,12288,0.01277511070171992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,3584,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,4096,0.012411555482281579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,3072,0.006309333360857434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,12288,0.02630666726165348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,2560,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,3072,0.010440000229411654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,2048,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,2560,0.009354666703277165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,1536,0.004553777890072929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,4096,0.00793688909875022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,1024,0.003910222401221593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,6144,0.01585155559910668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,2048,0.008674666285514832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,768,0.003540444291300244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,512,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,16384,0.03307911091380649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,256,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,1024,0.006656888872385025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,768,0.006206222292449739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,512,0.005629333357016246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,64,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,256,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,32,0.002540444541308615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,64,0.004936000125275718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,5120,0.014162666267818876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,32,0.004946666873163647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,128,0.002936888900068071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,51200,0.026698667142126296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,128,0.005075555708673266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,1536,0.007668444679843054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,3584,0.011437333292431302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,16384,1.5846195220947266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,16384,1.7942186991373699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,16384,0.5171439912584093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,51200,1.807079103257921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,12288,1.1184551450941298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,51200,4.921707577175565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,12288,1.2650906244913738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,16384,1.4971307118733723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,10240,1.0040409300062392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,10240,0.32853155665927464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,12288,0.40239111582438153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,51200,6.10588879055447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,8192,0.7225119802686902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,12288,1.1223760181003146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,10240,1.0626950793796115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,8192,0.27320888307359487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,8192,0.8303750885857476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,10240,0.8733306460910373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,7168,0.6860328780280219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,7168,0.23824710316128203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,7168,0.7134559949239095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,6144,0.5753128793504503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,6144,0.21180533038245308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,8192,0.7357795503404406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,5120,0.45284091101752383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,7168,0.6409280035230849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,5120,0.18539644612206352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,5120,0.5268026457892524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,4096,0.4069831106397841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,6144,0.5546942287021214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,4096,0.41659998893737793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,4096,0.1499235497580634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,3584,0.3231644365522596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,3584,0.13422755400339761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,3584,0.3597413433922662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,3584,0.32549601131015354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,3072,0.3006977770063612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,3072,0.1185377836227417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,3072,0.29177512062920463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,5120,0.45493512683444554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,3072,0.28151199552747935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,4096,0.38159910837809247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,2560,0.2583848900265164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,2048,0.2038231160905626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,2560,0.24786578284369576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,1536,0.14372355408138698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,2048,0.21122666200002035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,2048,0.09003111388948228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,2048,0.19928177197774252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,2560,0.22573865784539116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,1536,0.07166044579611884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,1536,0.16733955012427437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,2560,0.10537778006659614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,6144,0.6271226671006944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,768,0.0888764460881551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,51200,5.017788357204862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,1024,0.05714311202367147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,512,0.06051911248101128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,768,0.10248266988330418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,768,0.05077155431111654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,768,0.09765955474641587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,256,0.049159109592437744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,1024,0.11418222056494819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,512,0.08758755524953206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,512,0.08091911342408922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,128,0.03768444392416213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,256,0.07971644401550293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,256,0.06393955813513862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,1024,0.12632444169786242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,1024,0.11853333314259847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,64,0.03512710995144314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,128,0.07396622498830159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,32,0.03644266724586487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,128,0.05099911159939236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,512,0.04565155506134033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,32,0.07437066899405585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,1536,0.16240533192952475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,128,0.04079733292261759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,64,0.07377866903940837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,256,0.04220266805754768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,65536,1.8981439802381728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,51200,1.4223386976453993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,51200,3.7826961941189237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,65536,4.840534210205078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,16384,1.3851075702243383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,16384,0.39849421713087296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,51200,4.7319793701171875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,16384,1.4965573416815863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,12288,1.0549137327406142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,65536,6.553738488091363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,12288,0.3170986705356174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,10240,0.7221066686842176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,12288,0.9822177886962891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,16384,1.1364249123467338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,10240,0.8429839875963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,12288,0.8436249097188314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,8192,0.6811875767178006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,8192,0.6524542172749838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,8192,0.23606843418545195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,7168,0.49910487069023984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,10240,0.26042400466071236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,8192,0.5562399758232964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,7168,0.5646417935689291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,51200,3.919714609781901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,7168,0.18997244040171304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,10240,0.7157964706420898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,6144,0.4616702397664388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,6144,0.1658524407280816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,7168,0.5094524489508735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,65536,5.084649827745226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,5120,0.45064444012112087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,4096,0.3518657684326172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,5120,0.35576889250013566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,4096,0.3172444502512614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,3584,0.2523075474633111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,4096,0.13352089458041722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,5120,0.35390400886535645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,4096,0.29263822237650555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,3072,0.22864445050557455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,5120,0.14608089129130045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,3072,0.09721955325868394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,3072,0.23756443129645455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,2560,0.20269244247012666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,6144,0.4717751079135471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,2560,0.20278933313157824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,2560,0.08255822128719754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,3584,0.10520711210038926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,6144,0.43393511242336696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,2560,0.19774756166670057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,2048,0.07167110840479533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,3584,0.2568275663587782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,2048,0.15729244550069174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,3072,0.23031732771131727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,1536,0.13221689065297446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,1024,0.09123822053273518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,1536,0.12491022215949164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,3584,0.2744435469309489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,1024,0.09967377450731064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,768,0.06632266441980998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,1024,0.09547289212544759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,2048,0.1669600009918213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,512,0.048807111051347524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,768,0.081185777982076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,768,0.041386667225095965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,768,0.07939466502931383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,1536,0.05611199802822537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,256,0.03625333309173584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,512,0.0702400008837382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,512,0.065519114335378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,128,0.03064089020093282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,2048,0.1660533348719279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,256,0.06356000237994723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,256,0.05178222060203552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,64,0.026016000244352553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,128,0.03365955419010586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,128,0.05891199906667074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,32,0.026371555195914373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,128,0.04183822207980686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,64,0.05894044372770521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,32,0.05942666530609131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,1024,0.04550933175616794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,1536,0.13484266069200304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,65536,0.5923031171162924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,65536,1.5679510964287653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,512,0.03758488761054145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,51200,1.2118381924099393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,256,0.034930666287740074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,65536,1.9530853695339625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,16384,0.1406924459669325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,51200,1.4552977879842122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,12288,0.30013155937194824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,16384,0.3644355667961969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,16384,0.37290488349066836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,12288,0.30902134047614205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,65536,1.652507570054796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,10240,0.2545582188500298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,12288,0.27651911311679417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,8192,0.23874489466349283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,16384,0.4433297846052382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,51200,1.2610587014092338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,12288,0.107951111263699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,8192,0.07923466629452176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,8192,0.187208890914917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,7168,0.21894578138987222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,51200,0.4573129018147786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,7168,0.188554671075609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,6144,0.18374133110046387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,10240,0.2569039927588569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,10240,0.235335111618042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,6144,0.060714668697781034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,6144,0.1578728887769911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,5120,0.1495528883404202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,8192,0.25651733080546063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,4096,0.1145697832107544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,6144,0.14417244328392878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,5120,0.13209688663482666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,5120,0.12403910689883763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,7168,0.06646222538418241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,10240,0.09400800201627944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,4096,0.04353244437111748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,4096,0.10264177454842462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,3584,0.038615110847685076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,3584,0.09045688973532783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,5120,0.05055200060208639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,3072,0.08271021975411309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,4096,0.10668622122870551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,7168,0.16528977288140193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,2560,0.07103555732303195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,3584,0.09439377652274238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,2560,0.07168710894054837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,3072,0.09546399778789944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,2048,0.057131555345323354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,2560,0.030037333567937214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,2048,0.026047999660174053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,2048,0.05941422118080986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,1536,0.04609955681694878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,2048,0.059345775180392794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,3584,0.10261778036753337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,3072,0.033815112378862165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,1536,0.021135111649831135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,1536,0.04758222235573662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,1536,0.048560000128216214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,768,0.02537422213289473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,1024,0.03660622239112854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,2560,0.06939822435379028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,512,0.019366222951147292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,768,0.02940800123744541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,768,0.0326533317565918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,512,0.024480889240900677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,256,0.014535999960369535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,512,0.02777066661251916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,1024,0.017840888765123155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,256,0.02200711104604933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,128,0.011892444557613797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,1024,0.03846577803293864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,768,0.01620533400111728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,128,0.013803555733627744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,64,0.010472888747851053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,512,0.015492444237073263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,3072,0.08080799712075128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,64,0.020959110723601446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,32,0.020963556236690942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,256,0.02149244480662876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,128,0.021184888150956895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,128,0.01853511068556044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,1024,0.03244444396760728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,65536,1.4064142439100478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,65536,0.4555315441555447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,256,0.013914666242069669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,51200,0.928259531656901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,65536,1.556926197475857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,51200,0.35289067692226833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,51200,1.0940862231784396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,16384,0.32967911826239693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,12288,0.2207191122902764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,16384,0.29468711217244464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,65536,1.2455404069688585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,12288,0.08494044674767388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,12288,0.23345600234137642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,16384,0.30361776881747776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,12288,0.21386933326721191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,32,0.010074666804737514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,10240,0.0739591121673584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,10240,0.20317422019110787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,10240,0.18362311522165933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,16384,0.11082310809029473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,8192,0.059120886855655246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,7168,0.17298844125535753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,8192,0.14850221739874944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,7168,0.05396622088220385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,7168,0.13978311750623915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,10240,0.2434879938761393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,7168,0.13001155853271484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,6144,0.04769511024157206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,51200,0.8905306922064887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,6144,0.12081066767374675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,8192,0.16308711634741888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,4096,0.0900506642129686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,5120,0.10216355323791504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,8192,0.15827643871307373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,5120,0.09876977735095555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,3584,0.07973511351479425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,4096,0.0341288877858056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,6144,0.1368488868077596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,4096,0.08156177732679579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,3584,0.02980355421702067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,3584,0.07377688752280341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,3584,0.07306844658321805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,6144,0.1134622229470147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,2560,0.05792178048027886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,3072,0.0644328859117296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,3072,0.02730133301681942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,3072,0.06572533316082425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,2048,0.04560888806978861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,2560,0.05579377545250786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,4096,0.08306222491794162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,2560,0.056292441156175405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,2048,0.04698044392797682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,3072,0.07154399818844266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,1536,0.03534311056137085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,2048,0.0481315553188324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,5120,0.04000622365209792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,1024,0.025847110483381484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,1536,0.040596445401509605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,2560,0.024711999628278945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,1024,0.028716444969177246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,1024,0.01521511044767168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,768,0.02055288851261139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,2048,0.021356445219781663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,5120,0.11208621660868327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,512,0.018605333235528734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,768,0.013914666242069669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,768,0.026962666047943964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,1536,0.017619555195172627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,256,0.014853333433469137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,512,0.013008000122176277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,512,0.02290133304066128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,128,0.012655999925401477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,256,0.01754844519827101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,256,0.018068444397714403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,1024,0.03234222200181749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,64,0.011752888560295105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,128,0.01721955504682329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,128,0.012098666694429187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,32,0.011741333537631564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,768,0.02370577719476488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,1536,0.037392887804243304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,64,0.016864000095261466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,512,0.019985778464211356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,256,0.012080888781282636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,65536,0.3730124367607965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,65536,0.9914106792873807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,51200,0.7541795306735568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,128,0.01664177742269304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,65536,1.1692470974392362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,16384,0.2568746672736274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,51200,0.29943821165296763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,16384,0.09212533632914226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,16384,0.26993245548672146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,12288,0.2091777722040812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,16384,0.2437288761138916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,65536,1.0499821768866646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,12288,0.07295644283294678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,10240,0.15185956160227457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,51200,0.7846399943033854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,10240,0.06193066967858208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,8192,0.1368231111102634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,10240,0.15766578250461155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,8192,0.05094933178689745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,8192,0.13411022557152644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,8192,0.13148177994622126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,12288,0.2014862166510688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,7168,0.11828800042470296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,32,0.017124444246292114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,7168,0.04487288991610209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,6144,0.08843377563688491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,51200,0.9381244447496203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,6144,0.039622220728132464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,6144,0.09928977489471436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,7168,0.1241004467010498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,5120,0.08719111151165432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,12288,0.18393777476416695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,5120,0.08556889163123237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,7168,0.11476000150044759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,10240,0.16830755604637992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,3584,0.06545155578189425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,4096,0.02944355540805393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,4096,0.0712897777557373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,5120,0.09189955393473308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,4096,0.07121688789791532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,3072,0.057892441749572754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,3584,0.02630399995379978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,3584,0.06280710962083605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,2560,0.04773777723312378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,3072,0.023863111933072407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,3072,0.055514666769239635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,3072,0.056757330894470215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,4096,0.07078044282065497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,2048,0.03700622253947788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,2560,0.04997511042488945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,2048,0.01920000049802992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,6144,0.10303466849856907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,2048,0.04251022140185038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,3584,0.06547555658552381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,1536,0.0326275560590956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,5120,0.034104890293545194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,1536,0.035184890031814575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,1536,0.016688888271649677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,2560,0.02199466692076789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,2560,0.047741333643595375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,768,0.015600888265503777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,1024,0.025133333272404138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,1024,0.028466665082507666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,768,0.012786666552225748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,768,0.02066044509410858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,1536,0.0274311105410258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,768,0.02423200011253357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,512,0.011966221862369113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,256,0.01070577816830741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,1024,0.017822222577200997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,1024,0.014176888598336114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,256,0.016653334101041157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,128,0.009338666995366415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,512,0.014269333746698169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,128,0.01516177753607432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,512,0.017394666870435078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,128,0.01566133399804433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,64,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,512,0.021073778470357258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,32,0.008927110996511247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,64,0.01499022212293413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,32,0.014886221951908536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,256,0.011196444431940714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,2048,0.040277332067489624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,65536,0.8299199740091959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,65536,0.32082756360371906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,51200,0.6288480228847927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,65536,0.935431162516276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,256,0.015609777635998197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,16384,0.21570132838355172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,51200,0.24848445256551108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,51200,0.7556604279412164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,12288,0.17889689074622261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,16384,0.08142933580610487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,16384,0.21606932746039498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,65536,0.8296026653713651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,10240,0.16730044947730172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,128,0.011417778001891242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,51200,0.6650355656941732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,10240,0.05273600088225471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,10240,0.14171022839016384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,10240,0.13287111123402914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,8192,0.11065066523022121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,8192,0.11007999711566502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,7168,0.11293244361877441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,12288,0.062260442309909396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,12288,0.1639066669676039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,12288,0.15735822253757054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,6144,0.0959048867225647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,7168,0.09706488582823013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,8192,0.12733244895935059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,5120,0.0825351079305013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,6144,0.03383377856678433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,6144,0.08564177486631606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,6144,0.08468088838789199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,8192,0.04299644298023648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,4096,0.06299466556972928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,5120,0.07212089167700873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,16384,0.20572266313764784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,5120,0.07334578037261963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,3584,0.056495997640821666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,4096,0.025050666597154405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,4096,0.05972533093558418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,7168,0.09886488649580215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,7168,0.03803199860784743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,3584,0.052891555759641856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,3584,0.05530666642718845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,2560,0.03966755668322245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,3072,0.04679110977384779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,3072,0.04953333404329088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,5120,0.03011911114056905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,2560,0.04030133287111918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,2560,0.04308088951640659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,3072,0.05049688948525322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,3584,0.022879110442267522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,1536,0.025822222232818604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,2048,0.037422223223580256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,2048,0.01722311145729489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,3072,0.021017778250906203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,1024,0.017575111654069688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,1536,0.014163555370436775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,1536,0.027994665834638808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,1536,0.031201776531007554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,2560,0.01953866746690538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,1024,0.021410667233996924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,1024,0.02525955604182349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,4096,0.06115466356277466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,768,0.017847110827763874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,512,0.011408888631396823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,768,0.021202666891945735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,512,0.011060444845093621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,512,0.015207111835479736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,512,0.018087999688254464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,768,0.014551111393504672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,256,0.013528889252079858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,128,0.00737955586777793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,256,0.010384000009960597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,256,0.015282667345470853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,2048,0.03415022293726603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,768,0.01180266671710544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,64,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,128,0.01038933296998342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,32,0.008683555656009251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,128,0.01313244468635983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,128,0.013578666581047906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,64,0.012823111481136747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,32,0.01293333371480306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,2048,0.0329039990901947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,65536,0.7215110990736219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,65536,0.27117599381340873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,65536,0.8293875588311089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,51200,0.5709813435872396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,256,0.008365333080291748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,51200,0.2098933325873481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,51200,0.6913919978671603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,65536,0.7032195727030436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,1024,0.012503999802801343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,16384,0.06678755415810479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,51200,0.5501306851704916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,16384,0.17579910490247938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,12288,0.1443448861440023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,12288,0.05361955695682102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,12288,0.1331040064493815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,16384,0.18765689267052543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,10240,0.12195555369059245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,10240,0.11373421880933975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,12288,0.14242843786875406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,8192,0.09887733062108357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,7168,0.0886453323894077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,8192,0.09326489104164971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,10240,0.10696178012424046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,7168,0.08771733442942302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,7168,0.03279733326699998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,7168,0.08233600192599826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,8192,0.09683022234174941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,6144,0.029189334975348577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,6144,0.07642755243513319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,6144,0.0727635555797153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,16384,0.19383377499050566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,4096,0.04842488964398702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,8192,0.03601422243648105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,5120,0.06241599718729655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,4096,0.05372533202171326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,6144,0.07112266620000203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,3584,0.044158223602506846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,10240,0.045143110884560474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,5120,0.061667554908328585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,3584,0.04815999997986687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,3584,0.02011999984582265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,3584,0.047614223427242704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,5120,0.06508177518844604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,2560,0.03293155630429586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,3072,0.04239111145337423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,4096,0.02197066611713833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,4096,0.052833778990639575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,2560,0.036870221296946205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,2048,0.025441777375009324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,2560,0.017162667380438913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,2560,0.03745333353678385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,3072,0.037500444385740496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,1536,0.020195555355813768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,2048,0.032602667808532715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,2048,0.015225778023401896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,5120,0.02609511050913069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,1536,0.025807999902301367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,1536,0.027294221851560805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,1024,0.019126221537590027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,768,0.012351999680201212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,1024,0.022600889205932617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,1024,0.011643555429246692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,768,0.01074222226937612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,768,0.016275554895401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,768,0.019170666734377544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,2048,0.031090666850407917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,3072,0.018560889694425795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,256,0.007287999822033777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,512,0.010092444717884064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,512,0.016192000773217943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,256,0.012416889270146688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,256,0.00960888879166709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,128,0.006953777538405524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,512,0.009520000053776635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,128,0.012103111379676394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,64,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,128,0.013464888764752282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,128,0.009718221922715506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,1024,0.01513155632548862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,3072,0.04259555538495382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,512,0.013825777504179211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,64,0.012084444363911947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,32,0.01203555530971951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,256,0.013119999733236102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,65536,0.7085919910007052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,51200,0.5376737912495931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,65536,0.730012469821506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,32,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,1536,0.01314933349688848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,16384,0.1673440006044176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,51200,0.19746222760942247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,65536,0.6907626787821451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,12288,0.12921067078908285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,16384,0.06279999679989286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,16384,0.17040888468424478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,16384,0.16933155059814453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,65536,0.24902134471469453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,12288,0.12759733200073242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,12288,0.13003467188941106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,10240,0.10759111245473225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,10240,0.04295377930005392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,8192,0.08110755681991577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,10240,0.10994666814804077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,8192,0.03476266728507148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,8192,0.08732000324461196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,7168,0.07160622543758817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,8192,0.09166400300131904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,51200,0.5378524462381998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,7168,0.03112799922625224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,7168,0.0802382230758667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,51200,0.5780284669664171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,6144,0.028151111470328435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,5120,0.06046399805280897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,12288,0.04976800084114075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,6144,0.07166133324305217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,5120,0.057018664148118764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,4096,0.04738577869203356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,6144,0.0673688915040758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,4096,0.046514666742748685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,3584,0.04129066732194688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,4096,0.052667554881837636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,10240,0.11822756131490071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,3584,0.019779556327395968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,3584,0.041562666495641075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,3072,0.03639111253950331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,3584,0.047460443443722196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,5120,0.02486755616135067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,2560,0.03087555699878269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,3072,0.01824355622132619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,3072,0.041730665498309664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,4096,0.021615111165576514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,2560,0.03209599852561951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,2560,0.01664888858795166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,2560,0.037400000625186496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,2048,0.014827556080288358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,1536,0.020792000823550753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,5120,0.06193511353598701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,3072,0.03664177656173706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,1536,0.022675555613305833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,1024,0.015495111544926962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,6144,0.07077155510584514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,1536,0.026644445127911035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,2048,0.02497333288192749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,768,0.012463111016485425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,1024,0.011420444481902652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,1024,0.02104088829623328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,768,0.014819555812411837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,512,0.009642666412724389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,768,0.018594667315483093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,512,0.012527111503813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,1536,0.013077333569526672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,512,0.015959110524919298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,512,0.00980977796845966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,1024,0.017198221551047433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,2048,0.027424888478385076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,256,0.009343999955389235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,2048,0.03138666682773166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,768,0.010458666417333813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,128,0.009380444056457942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,128,0.011056000159846412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,64,0.0064737776087390045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,128,0.012436444560686747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,256,0.007729777859316931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,64,0.011075555450386472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,32,0.011062222222487131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,7168,0.07735200060738458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,128,0.006800888727108638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,256,0.013097777962684631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,65536,0.2317662239074707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,65536,0.6215582423739964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,32,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,16384,0.12000444200303818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,51200,0.4826141993204753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,65536,0.5289831161499023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,256,0.01146755533085929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,65536,0.47449154324001735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,51200,0.4110479884677463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,16384,0.13552355766296387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,51200,0.4418453375498454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,12288,0.10883200168609619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,10240,0.10074933369954427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,12288,0.10702133178710938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,51200,0.18545511033799914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,10240,0.038663112454944186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,8192,0.07438755697674222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,16384,0.058463997311062284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,16384,0.14520977603064641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,12288,0.09360533290439182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,8192,0.03155111273129781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,8192,0.07538577583101061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,8192,0.07589422331915961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,6144,0.05434133277999031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,7168,0.06650577651129828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,7168,0.06606311268276639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,6144,0.058202664057413735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,10240,0.09190666675567627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,10240,0.08942488829294841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,7168,0.05496799945831299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,5120,0.04918577935960558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,12288,0.04606311188803779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,5120,0.02211911148495144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,4096,0.037929776642057635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,5120,0.05017422305213081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,4096,0.01942488882276747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,5120,0.04550666610399882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,6144,0.058371557129753955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,3584,0.03689422210057577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,3584,0.03902311126391093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,3584,0.01789066692193349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,3072,0.026846221751636926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,6144,0.025399110383457605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,3072,0.03263466556866964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,3584,0.0347306662135654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,3072,0.03498044278886583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,3072,0.01633066601223416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,2560,0.014888889259762235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,2560,0.0284222231970893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,2048,0.02071199980047014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,2560,0.031038221385743883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,7168,0.02759644389152527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,1536,0.01607644392384423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,2048,0.013968000809351603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,2048,0.02416177756256527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,1024,0.012074666718641916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,1536,0.011756444142924415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,1536,0.019744000501102872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,1536,0.022376888328128394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,768,0.010093332992659675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,1024,0.018223111828168232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,768,0.013215999636385175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,768,0.015961777832772996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,512,0.008036444584528605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,2048,0.026590221458011206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,4096,0.0406897763411204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,4096,0.04263200031386482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,512,0.009069333473841349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,512,0.01443733274936676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,256,0.010426667001512317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,256,0.01237511138121287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,256,0.008647110727098253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,128,0.010072888599501716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,1024,0.010040000081062317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,128,0.009006222089131674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,64,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,128,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,32,0.005611555443869696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,64,0.010070222119490305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,2560,0.02418488926357693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,32,0.010079111489984725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,1024,0.01531733406914605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,65536,0.4921021991305881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,128,0.012072000238630505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,65536,0.1890151103337606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,51200,0.38288267453511554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,65536,0.5067191123962402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,16384,0.13052977455986872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,512,0.011405333048767514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,51200,0.1499368879530165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,256,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,12288,0.08909510903888279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,51200,0.40538933542039657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,16384,0.13423200448354086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,768,0.009452444811662039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,12288,0.03917866614129808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,10240,0.06923111279805501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,65536,0.5151386790805393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,10240,0.07742844687567817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,8192,0.06171822547912598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,10240,0.08796977996826172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,8192,0.0635591083102756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,12288,0.09228266610039605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,8192,0.07298133108350965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,8192,0.02757688860098521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,51200,0.3958017826080322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,7168,0.05641688903172811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,7168,0.06472977664735582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,6144,0.039817776944902204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,7168,0.025240888198216755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,6144,0.022760000493791368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,6144,0.04912799927923414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,6144,0.05726933479309082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,7168,0.04529688755671183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,5120,0.04211555586920845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,5120,0.02037244372897678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,4096,0.03156088789304098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,10240,0.03394577900568644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,5120,0.04951644606060452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,16384,0.049581332339180835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,4096,0.017640888690948486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,4096,0.041611555549833514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,5120,0.04027022255791558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,3584,0.016319110989570618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,3072,0.024561777710914612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,3584,0.03797600004408095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,3072,0.015003555350833468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,3072,0.028226666980319556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,2560,0.020851555797788832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,3072,0.03410044312477112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,16384,0.12028177579243977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,4096,0.03551377852757772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,2560,0.024857777688238356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,2048,0.016785777277416654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,2560,0.013795555465751223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,2560,0.030138668086793687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,1536,0.01313777764638265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,2048,0.012435555458068848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,2048,0.020953777763578627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,12288,0.10285333130094741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,1024,0.009975999593734741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,1536,0.011114666859308878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,1536,0.017255999975734286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,1536,0.02177422245343526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,768,0.008612444831265343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,3584,0.02587022218439314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,1024,0.017702221870422363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,768,0.009335111412737105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,512,0.007290666302045186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,3584,0.031657778554492526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,768,0.015588444140222339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,2048,0.025609778033362493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,256,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,512,0.008720888859695857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,512,0.013456000222100152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,1024,0.013467555244763693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,128,0.005576000031497743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,256,0.01144977741771274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,768,0.01146399974822998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,64,0.005233777893914117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,128,0.008388444781303406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,32,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,128,0.01128000020980835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,64,0.009032888544930352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,32,0.009005332986513773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,65536,0.31910133361816406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,256,0.008382221890820397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,65536,0.44950400458441836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,128,0.009038222332795462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,65536,0.3776248825920953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,512,0.01038933296998342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,51200,0.3518222173055013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,1024,0.009944000177913243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,51200,0.29855110910203725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,256,0.009349333743254343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,12288,0.06926400131649442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,16384,0.039348443349202476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,65536,0.14889332983228895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,16384,0.10063199864493476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,10240,0.06161955330106947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,12288,0.031261334816614784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,12288,0.07761955261230469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,51200,0.11958489153120254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,10240,0.027405333187845018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,8192,0.05115911033418444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,10240,0.0675519969728258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,8192,0.022338666849666174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,16384,0.10832177930408055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,51200,0.2975920041402181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,8192,0.0559555557039049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,12288,0.08397066593170166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,7168,0.051775111092461475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,7168,0.050195554892222084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,16384,0.10010755724377102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,6144,0.0456026660071479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,6144,0.018505778577592637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,5120,0.03123733401298523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,6144,0.04436622063318888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,8192,0.0583546691470676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,7168,0.04000533289379544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,4096,0.027313777142100867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,5120,0.03870577613512675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,5120,0.03923555546336704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,4096,0.014716444744004143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,3584,0.022879999544885423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,4096,0.03291733397377862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,10240,0.07034666670693292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,3072,0.020558221472634208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,3584,0.02926666537920634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,3584,0.030701332622104224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,3072,0.012776888906955719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,5120,0.01688888834582435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,3072,0.027264888087908428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,7168,0.020266667008399963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,6144,0.03908533189031813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,4096,0.032814221249686346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,2560,0.011436444189813403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,2560,0.024294222394625347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,3584,0.013571555415789286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,2048,0.019307555423842538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,2048,0.021077333225144282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,3072,0.026163554853863184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,2560,0.01799199978510539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,1536,0.015969778100649517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,1024,0.009813333551088968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,1536,0.009696889254781935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,1536,0.01791111131509145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,768,0.008159110943476358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,1024,0.012544000314341651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,1024,0.014937778313954672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,2048,0.010703110860453712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,1536,0.012415111064910889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,512,0.006980444822046492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,768,0.008604444563388824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,512,0.009673777553770278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,512,0.012087110843923358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,512,0.008285333712895712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,256,0.005586666779385672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,256,0.008738666772842407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,256,0.010372444159454769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,128,0.00489688871635331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,256,0.007700444095664554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,128,0.008662222160233391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,128,0.010390222072601318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,1024,0.009001777403884465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,128,0.007820444802443186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,32,0.004577777865860197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,768,0.011057777537239922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,32,0.008572444319725037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,768,0.01314133322901196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,65536,0.36075199974907773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,65536,0.14169243971506754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,51200,0.2798408932156033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,2048,0.01496088835928175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,65536,0.3682088851928711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,51200,0.11194933785332574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,16384,0.0904266635576884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,51200,0.31303199132283527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,51200,0.2939457893371582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,12288,0.06902844376034208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,16384,0.10142933660083348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,2560,0.02267644471592373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,12288,0.0753199987941318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,10240,0.05746489100986057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,12288,0.03062844607565138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,64,0.008620444271299574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,10240,0.06394844584994845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,10240,0.02719466719362471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,8192,0.04602488875389099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,10240,0.06587555673387316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,8192,0.022043555974960327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,8192,0.0521964430809021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,7168,0.04127466678619385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,65536,0.4053048822614882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,8192,0.055213332176208496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,6144,0.03460977805985345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,7168,0.020276443825827706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,12288,0.07908355527453952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,7168,0.04643911123275757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,5120,0.030258665482203167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,6144,0.0185244447655148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,6144,0.0409786668088701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,6144,0.04351466563012865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,16384,0.0378115541405148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,4096,0.024872889121373493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,5120,0.035367111365000405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,5120,0.039000888665517174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,3584,0.022687110635969374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,64,0.004572444491916233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,4096,0.03004266818364461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,4096,0.0324853327539232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,3072,0.0200311111079322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,3584,0.013516444298956128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,3584,0.02703999976317088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,7168,0.04946488804287381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,2560,0.017650667164060805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,3072,0.012429333395428128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,3072,0.023759111762046814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,16384,0.09917689032024807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,2560,0.011510222322411008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,2560,0.020615110794703167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,2560,0.02426044477356805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,4096,0.014510222607188754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,1536,0.01258044441541036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,2048,0.017858665850427415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,2048,0.02035999960369534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,1536,0.014855111638704935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,1536,0.009708444277445475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,1024,0.00923999978436364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,3072,0.026799110902680293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,2048,0.014961777461899651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,768,0.008016889293988546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,1024,0.014583110809326172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,1024,0.009005332986513773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,2048,0.010394666757848527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,512,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,768,0.010379555324713389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,768,0.013085333009560903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,3584,0.029325332906511094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,1536,0.01754666699303521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,512,0.009134222235944536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,512,0.008039111064540016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,512,0.012140444583363004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,128,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,256,0.007689778175618913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,256,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,1024,0.011753777662913004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,128,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,128,0.007977777885066139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,128,0.010405333505736457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,32,0.005586666779385672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,64,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,256,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,32,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,65536,0.3049573368496365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,5120,0.01645066671901279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,65536,0.1411857737435235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,51200,0.23307198948330352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,64,0.004562666846646203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,768,0.008378666308191087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,65536,0.36727110544840497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,16384,0.06859911150402494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,51200,0.27806221114264595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,51200,0.28926489088270396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,16384,0.03644888930850559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,16384,0.0866231123606364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,16384,0.09924266735712688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,256,0.010219555762079027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,12288,0.06644622484842937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,10240,0.047451555728912354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,12288,0.07646577888064914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,10240,0.026542221506436665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,10240,0.05679111348258125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,10240,0.06684177451663546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,12288,0.058097779750823975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,8192,0.021664000219768945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,8192,0.046428445312711925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,8192,0.05470755696296692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,7168,0.01971733404530419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,7168,0.04219377703136868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,6144,0.029129776689741347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,7168,0.048615111245049365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,51200,0.1094391081068251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,6144,0.01806400054030948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,8192,0.040392001469930015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,6144,0.043142222695880465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,5120,0.03198844525549147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,7168,0.03330311179161072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,4096,0.02083288960986667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,12288,0.029834667841593426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,4096,0.026770666241645813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,3584,0.01845066746075948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,5120,0.02405422263675266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,3584,0.024263999528355066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,5120,0.01611288885275523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,3584,0.013071111506885953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,3072,0.01662400033738878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,65536,0.3569635550181071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,3072,0.0214684440029992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,2560,0.014523555835088095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,6144,0.037230223417282104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,2560,0.018592889110247295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,2560,0.011300444602966309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,2560,0.023694222172101338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,2048,0.0124177775449223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,2048,0.016140444411171805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,4096,0.03199111090766059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,1536,0.010290666586822933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,2048,0.010374222364690568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,1536,0.013448889056841532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,1536,0.01733688844574822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,5120,0.03753066725201077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,1536,0.009681777821646797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,1024,0.010774222513039907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,3072,0.02628622286849552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,768,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,1024,0.008781332936551835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,768,0.009583111438486312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,768,0.012892444100644855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,3584,0.02936089038848877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,768,0.00832799987660514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,512,0.00866844422287411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,4096,0.014067555467287699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,512,0.012103999654452005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,256,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,512,0.008008889026112026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,256,0.007638221813572778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,256,0.007688889073001013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,128,0.004286222159862518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,256,0.010382222632567087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,1024,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,64,0.003891555385457145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,128,0.007358222372002072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,32,0.004230222354332606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,3072,0.012151111331250934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,32,0.007326222128338284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,65536,0.20245333512624106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,512,0.005979555762476391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,65536,0.3081208864847819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,128,0.010391111175219217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,51200,0.1538106600443522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,65536,0.13216355111863878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,2048,0.020444444484180875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,16384,0.06040800280041165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,51200,0.10495644145541722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,51200,0.2428577740987142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,51200,0.2889048788282606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,16384,0.07635111278957792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,12288,0.04380888740221659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,16384,0.09793600108888413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,128,0.007687999970383114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,12288,0.05861510833104452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,65536,0.36619822184244794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,10240,0.03812888926929898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,1024,0.01477599971824222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,12288,0.07547911008199056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,8192,0.030643555853101943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,10240,0.025626666016048853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,10240,0.05016533202595181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,10240,0.06484888659583198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,8192,0.020626667473051283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,8192,0.041492445601357356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,8192,0.054442667298846774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,7168,0.03722933265897963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,12288,0.028396444188223943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,6144,0.02367288867632548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,64,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,6144,0.017533333765135873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,5120,0.02115733259254032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,6144,0.04235644472969902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,7168,0.026781333817376032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,5120,0.015507555670208402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,4096,0.016905777984195285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,5120,0.037068443165885076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,7168,0.019140443868107267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,4096,0.02370133333735996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,3584,0.015691555208630033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,4096,0.031054221921496924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,3584,0.012961777547995249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,6144,0.032905777295430504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,3072,0.014089778065681458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,16384,0.03500888744990031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,5120,0.028551111618677776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,2560,0.012418666647540199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,3072,0.011766222616036734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,4096,0.013449778159459433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,7168,0.04829955432150099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,2560,0.01686755485004849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,2560,0.011003555523024665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,2560,0.02279022170437707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,3584,0.021381333470344543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,3584,0.02880711025661892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,2048,0.014501333236694336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,2048,0.019648889700571697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,3072,0.01904977858066559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,1536,0.012248888611793518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,3072,0.02563200063175625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,1536,0.009497777455382878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,1024,0.008660444782839881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,1024,0.010042666561073726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,2048,0.010774222513039907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,1024,0.014527110589875115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,768,0.00832444429397583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,1536,0.009024000002278222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,512,0.005246222019195557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,768,0.012526222401195101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,512,0.008047999607192146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,256,0.004615111069546806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,768,0.005986666513813867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,256,0.007344000041484833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,256,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,768,0.008685333033402761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,128,0.003927999900446998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,256,0.010494222243626913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,128,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,1024,0.006952889263629913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,1536,0.01686311099264357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,32,0.004902222090297275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,64,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,32,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,65536,0.165702223777771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,2048,0.010048000348938836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,65536,0.08704088793860541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,65536,0.25925421714782715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,65536,0.21885422865549722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,51200,0.20144444041781953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,51200,0.17278756035698783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,51200,0.07001777489980061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,128,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,16384,0.06600355439715914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,51200,0.15509243806203207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,12288,0.03603377938270569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,16384,0.02502577834659153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,16384,0.062241779433356396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,12288,0.050796445873048574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,12288,0.020644444558355544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,10240,0.031883554326163396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,10240,0.04302755660480923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,64,0.004610666798220741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,512,0.012086222569147745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,10240,0.04235555397139656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,10240,0.018790221876568265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,8192,0.01497066683239407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,8192,0.03590755661328634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,7168,0.022422222627533808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,12288,0.04996088809437222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,7168,0.03243911266326904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,6144,0.019174221489164565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,512,0.00796088907453749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,128,0.010079111489984725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,6144,0.028431110911899145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,6144,0.029091556866963703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,8192,0.03591555688116286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,5120,0.016671111186345417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,16384,0.04874311221970452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,7168,0.013831111292044321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,7168,0.03226222263442145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,8192,0.024451555477248296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,4096,0.020663999848895602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,6144,0.012780444489585029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,3584,0.013034666577974955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,5120,0.024767110745112102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,3584,0.01867377758026123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,3584,0.010047111246320937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,3072,0.011843555503421359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,3584,0.019658666517999437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,4096,0.010663111176755695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,3072,0.016799110505315993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,3072,0.01789155602455139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,5120,0.011761777930789523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,4096,0.014988443917698331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,2048,0.009125333693292405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,2560,0.016407110624843173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,2048,0.013017777767446307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,2048,0.014523555835088095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,1536,0.007418666448858049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,2560,0.01070577816830741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,1536,0.011071110765139261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,5120,0.025591999292373657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,1024,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,2560,0.009296000003814697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,1024,0.00905866672595342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,4096,0.021539555655585393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,1024,0.01145155562294854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,2048,0.008831111093362173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,1024,0.007861333588759104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,768,0.007556444240940942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,512,0.004949333353175057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,768,0.010381333529949188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,1536,0.00831911133395301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,512,0.007339555356237624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,512,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,512,0.010040000081062317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,3072,0.009713778065310584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,768,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,256,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,128,0.0038782221575578055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,256,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,768,0.008016000191370646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,128,0.006680888848172293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,128,0.006455111006895701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,32,0.003896000070704354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,64,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,32,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,65536,0.1473697821299235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,256,0.008750222623348236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,65536,0.21463378270467123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,65536,0.21307199531131318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,65536,0.08557866679297553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,128,0.008648888932334052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,51200,0.06772533390257093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,16384,0.03733511103524102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,51200,0.17106400595770943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,16384,0.05485599901941088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,16384,0.06151466899447971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,16384,0.02365333338578542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,12288,0.026408000124825373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,256,0.004268444246715969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,12288,0.042479998535580106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,51200,0.1662666665183173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,10240,0.022097777989175584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,12288,0.01995822290579478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,10240,0.018240888913472492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,10240,0.03622666663593716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,8192,0.019056000643306308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,10240,0.04081244601143731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,2560,0.014839111102951897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,8192,0.014480888843536377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,8192,0.030208888981077407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,8192,0.034909334447648786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,7168,0.027319111757808264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,7168,0.01350488927629259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,6144,0.014852444330851236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,7168,0.031446221801969744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,6144,0.012459555433856117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,6144,0.02397955622937944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,6144,0.027269333600997925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,64,0.003919110943873723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,7168,0.016602666841612924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,5120,0.0207022229830424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,5120,0.011539555258221097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,5120,0.02392622166209751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,12288,0.04777955677774218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,4096,0.010416888528399997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,3584,0.010372444159454769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,4096,0.02088088956144121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,5120,0.01349422252840466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,3584,0.015969778100649517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,3584,0.01959111127588484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,1536,0.01294222225745519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,4096,0.01111644423670239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,3072,0.014476444986131458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,2560,0.008364444805516137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,3072,0.009402666654851701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,3072,0.01800622211562263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,4096,0.01759822169939677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,2048,0.007327111230956183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,2560,0.012787555654843649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,2560,0.01623911162217458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,3584,0.010077333284748925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,2048,0.011447110937701331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,2048,0.014290667242474027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,1536,0.009866666462686326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,1536,0.012823111481136747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,1536,0.00830488900343577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,2560,0.00906044410334693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,1024,0.008035555481910706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,3072,0.009685333404276106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,1536,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,1024,0.011409777734014722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,768,0.007339555356237624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,768,0.010409778190983666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,512,0.004271111140648524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,768,0.007365333537260692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,512,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,256,0.003842666745185852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,768,0.004699555536111196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,51200,0.1141528950797187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,256,0.006256888724035687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,128,0.003556444413132138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,256,0.008336000144481659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,1024,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,128,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,64,0.0032017777363459268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,128,0.008383110993438298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,512,0.007035555938879649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,32,0.003195555466744635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,512,0.009737778041097853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,2048,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,1024,0.007709333466158972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,64,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,32,0.006063999815119638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,51200,0.07971111271116468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,256,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,65536,0.08481511142518784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,65536,0.17405777507358125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,16384,0.031810667779710554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,51200,0.06856444146898058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,65536,0.2045226626925998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,12288,0.024373332659403484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,16384,0.02330488959948222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,128,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,16384,0.0575795571009318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,10240,0.020803555846214294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,12288,0.01956888867749108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,65536,0.1044071118036906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,12288,0.04502844479348925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,8192,0.017482666505707633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,10240,0.03087644444571601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,10240,0.03885866536034478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,8192,0.025796444879637823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,51200,0.1371271080440945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,7168,0.015834665960735746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,51200,0.1610142257478502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,7168,0.013413333230548434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,12288,0.03546488947338528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,6144,0.014470222923490735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,7168,0.0296382208665212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,10240,0.01759822169939677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,6144,0.012454222473833295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,6144,0.020558221472634208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,6144,0.026135999295446608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,8192,0.014534221755133735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,5120,0.018048889107174344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,8192,0.033349331882264875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,16384,0.045478221442964345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,4096,0.010383110907342698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,3584,0.009689778089523315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,4096,0.019682667321628995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,5120,0.012568888564904531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,3584,0.013904000322024027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,3072,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,3584,0.018404444058736164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,4096,0.0107004443804423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,5120,0.011617778076065911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,3072,0.009518221848540837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,3072,0.016918222109476726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,2560,0.0111191107167138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,2560,0.015259555644459195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,2048,0.00702044450574451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,5120,0.022960889670583937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,2048,0.010101333260536194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,2560,0.00795999997191959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,1536,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,4096,0.01517511076397366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,2048,0.008729777402347988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,2048,0.013804444836245643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,1024,0.004966222163703707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,1536,0.00869511150651508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,1536,0.012460444536474017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,2560,0.009072889056470659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,1024,0.007325333025720384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,768,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,3072,0.012506666282812754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,1024,0.01073155552148819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,7168,0.023165333602163527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,512,0.004272888931963178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,768,0.009869332942697737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,1536,0.008372444245550368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,512,0.00700800038046307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,512,0.006275555739800136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,256,0.003921777837806278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,1024,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,256,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,256,0.006683555742104848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,128,0.003593777616818746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,256,0.008719111482302347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,768,0.006685333533419504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,64,0.00332622230052948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,32,0.003573333223660787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,64,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,3584,0.009878222313192155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,65536,0.07184444533454047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,65536,0.16554755634731716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,768,0.007318221860461765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,65536,0.08482666810353597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,128,0.006666666517655055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,128,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,512,0.00941422250535753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,16384,0.024860444996092055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,51200,0.15965599483913845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,16384,0.043895999590555825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,16384,0.056968887646993004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,51200,0.05940888987647163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,16384,0.022849778334299724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,12288,0.034421334664026894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,32,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,12288,0.04435288906097412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,10240,0.018519111805491976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,12288,0.018927999668651156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,10240,0.017488888568348356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,65536,0.20168266031477186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,10240,0.03830666674507989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,128,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,8192,0.02439111140039232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,8192,0.014128888646761576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,8192,0.03223199976815118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,51200,0.13055288791656494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,51200,0.06833599673377143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,6144,0.01183733344078064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,7168,0.013117333253224691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,7168,0.028802666399214003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,10240,0.029527111185921565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,8192,0.014529777897728814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,6144,0.019747555255889893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,6144,0.025157334076033697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,12288,0.020376889242066275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,4096,0.01070577816830741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,5120,0.011354666617181567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,7168,0.012947555217478009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,4096,0.014862222803963555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,4096,0.010031999813185798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,3584,0.008001777860853408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,4096,0.019574221637513902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,3584,0.009723555710580613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,3584,0.013735111388895245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,3584,0.01815822223822276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,5120,0.017258667283587985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,5120,0.022703111171722412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,2560,0.006780444333950679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,7168,0.022256000174416438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,2560,0.01106844428512785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,3072,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,2048,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,2560,0.015062222878138224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,3072,0.012418666647540199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,2048,0.010071111222108206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,3072,0.016568889220555622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,1536,0.005151999907361137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,3072,0.009348444640636444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,1536,0.008026666939258575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,1536,0.00871733327706655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,1536,0.0120666664507654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,5120,0.010429333481523724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,2560,0.008852444589138031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,1024,0.007293333609898885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,1024,0.007663110891977946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,1024,0.01075644459989336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,768,0.007327111230956183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,2048,0.013587555951542325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,512,0.003563555578390757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,1024,0.004219555606444676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,512,0.006166222194830577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,256,0.003540444291300244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,768,0.004271999829345279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,256,0.005580444302823808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,768,0.006654222392373615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,256,0.006613333192136552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,128,0.0032417778339650896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,128,0.005316444569163852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,512,0.009039999710188972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,6144,0.012151111331250934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,64,0.0028684443483750024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,128,0.006610666712125142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,32,0.0031715554909573663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,128,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,65536,0.04498044318623013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,2048,0.008355555435021719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,65536,0.12489155928293864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,512,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,65536,0.08426577515072292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,51200,0.03763022356563144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,64,0.0053093334039052325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,51200,0.09835822052425808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,16384,0.015609777635998197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,51200,0.0676106678114997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,16384,0.03398488958676656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,65536,0.20077421930101183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,16384,0.02197244432237413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,12288,0.01441333360142178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,12288,0.01869244376818339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,12288,0.026522666215896606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,10240,0.01348533315791024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,256,0.00868444475862715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,10240,0.017270222306251526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,8192,0.011967110964987012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,16384,0.05616355604595608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,8192,0.019645333290100098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,8192,0.03159733282195197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,8192,0.01386666629049513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,51200,0.15841600630018446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,10240,0.023056889573733013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,7168,0.01812088820669386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,6144,0.010408889088365767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,7168,0.012990222209029727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,6144,0.016224000189039443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,6144,0.02490577763981289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,5120,0.009345778160625035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,32,0.005576888720194499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,5120,0.014141332772043018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,10240,0.03772355450524224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,5120,0.022023111581802368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,7168,0.02847466535038418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,5120,0.011074666347768573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,4096,0.010036444498433007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,4096,0.012444444828563266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,3584,0.007631999750932057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,4096,0.019136000010702345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,6144,0.012120889292822944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,3584,0.009711110757456886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,3072,0.006979555719428592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,3584,0.017532444662517972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,3072,0.009376000199053023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,4096,0.007975999679830339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,2560,0.006265777680608962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,3072,0.0162435554795795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,2560,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,2560,0.00962577760219574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,2048,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,2560,0.014815110299322339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,768,0.010076444182131026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,2048,0.008561777571837107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,2048,0.00870222184393141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,2048,0.01350933313369751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,3072,0.010390222072601318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,1536,0.008013333711359236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,1536,0.012129777835475074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,1024,0.006680888848172293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,1024,0.010774222513039907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,768,0.003940444439649582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,3584,0.011420444481902652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,768,0.006234666539563074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,1536,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,768,0.0094355551732911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,512,0.003579555700222651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,768,0.007344000041484833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,512,0.0069875551594628235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,512,0.005655999812814925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,512,0.009189333352777693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,7168,0.01146666705608368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,1536,0.004888888862397936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,256,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,256,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,1024,0.004239999999602635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,128,0.00508088908261723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,64,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,128,0.006659555352396435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,128,0.008160889148712158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,256,0.0032337777730491427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,1024,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,65536,0.035282668140199445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,128,0.0028702221396896574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,65536,0.08412000205781724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,51200,0.02962933315171136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,65536,0.20085867245992026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,32,0.0029057777590221832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,51200,0.09980177879333496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,51200,0.06743644343482123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,16384,0.013109332985348172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,51200,0.15840178065829807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,64,0.004943999979231092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,256,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,12288,0.011069333387745751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,16384,0.03337688909636603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,65536,0.12459111213684082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,16384,0.055810663435194224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,10240,0.010114666488435533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,12288,0.01868266695075565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,12288,0.04292444388071696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,10240,0.022946667340066697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,12288,0.04327733318010966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,10240,0.03679910964436001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,32,0.004931555440028508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,16384,0.021778666310840186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,8192,0.019482667247454327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,7168,0.00869422240389718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,8192,0.031621333625581526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,12288,0.026574222577942744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,6144,0.00832622249921163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,7168,0.0129120002190272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,7168,0.0281333327293396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,10240,0.01715555621518029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,8192,0.00905511114332411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,5120,0.00796622203456031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,6144,0.02478666603565216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,5120,0.011047999891969891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,5120,0.014177777700954013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,4096,0.007800000409285228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,5120,0.02200088898340861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,3584,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,4096,0.010083555347389644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,4096,0.012434666355450949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,4096,0.018946665856573317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,3072,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,3584,0.009713778065310584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,6144,0.015999999311235216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,3072,0.010538666612572141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,3072,0.009342222577995723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,3072,0.016192888220151264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,2560,0.009492444495360056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,2560,0.014831999937693277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,2560,0.008932444784376357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,2048,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,3584,0.011407111254003314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,6144,0.012109333442317115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,2048,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,2048,0.013471999930010902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,2560,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,8192,0.013863999810483722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,1024,0.00423200014564726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,1536,0.008014221986134848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,1536,0.01204533295498954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,7168,0.0181342214345932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,1024,0.006678222368160884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,1024,0.0076248885856734375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,1024,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,3584,0.017904000149832833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,768,0.007359999749395583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,1536,0.004586666822433472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,768,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,768,0.009680888719028896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,256,0.0032097777972618737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,512,0.005741333381997214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,1536,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,512,0.00904088881280687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,128,0.002942222274012036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,256,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,256,0.008018666671382057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,768,0.0035422220826148987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,64,0.0028124445428450904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,128,0.006649777707126405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,128,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,32,0.0028435554769304064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,512,0.0033653332955307434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,65536,0.03272711237271627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,64,0.005223999834722943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,512,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,32,0.004905777672926585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,16384,0.011863110793961419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,12288,0.010375111467308467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,65536,0.12435200479295518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,10240,0.00944266633854972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,16384,0.03290844294759963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,8192,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,12288,0.026144888665941026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,128,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,7168,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,10240,0.022788445154825848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,2048,0.008714666797055138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,8192,0.019279999865425956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,5120,0.008307555483447181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,7168,0.017660443981488545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,6144,0.015951111912727356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,3584,0.00703288863102595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,5120,0.014109333356221518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,3072,0.006313777632183499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,4096,0.012112889024946423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,256,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,2560,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,3584,0.0114720000161065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,3072,0.010388444695207808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,2560,0.009363555245929295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,6144,0.007640000018808577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,1536,0.004610666798220741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,1024,0.0038897775941424896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,2048,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,768,0.003584889074166616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,1536,0.007641777396202087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,512,0.00318755561278926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,1024,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,256,0.003015999992688497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,51200,0.09752088785171509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,128,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,512,0.005650666852792104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,64,0.0028702221396896574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,256,0.005245333330498801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,2048,0.00499555551343494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,32,0.002847111059559716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,65536,0.03165600034925673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,64,0.004933333231343163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,51200,0.026517333255873785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,65536,0.12615822421179876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,16384,0.009386666946940953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,768,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,12288,0.012311110893885294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,16384,0.0329022208849589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,10240,0.011112000379297467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,128,0.004938666605287128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,8192,0.009881777895821465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,7168,0.009027555584907532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,51200,0.027015111512607996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,10240,0.022675555613305833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,6144,0.009010666774378883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,8192,0.019242667489581637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,7168,0.0177484436167611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,4096,0.007885333564546373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,5120,0.014125333891974555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,3584,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,4096,0.01239111108912362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,3072,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,3584,0.011420444481902652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,2560,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,3072,0.010402667025725046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,2048,0.0052471111218134565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,2560,0.009363555245929295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,6144,0.01588800052801768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,1536,0.004577777865860197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,1024,0.0038933331767717996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,2048,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,768,0.0035582222044467926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,1536,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,512,0.0031973332580592898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,1024,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,256,0.002985777747299936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,768,0.006005333529578314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,128,0.002874666617976295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,512,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,64,0.0028497777465316984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,256,0.005283555636803309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,32,0.002530666689078013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,128,0.004957333207130432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,64,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,32,0.004947555561860402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,12288,0.025872000389628943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,4096,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,5120,0.008023111356629265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,51200,0.09781244728300308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,16384,1.278524398803711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,32,0.004965333475006951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,12288,0.8849324650234647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,16384,1.5080773035685222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,12288,0.37480444378323025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,10240,0.7872053252326118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,12288,1.0916417439778645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,16384,0.5029973453945583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,10240,0.31504888004726833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,10240,0.8842053413391113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,8192,0.6127697626749674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,8192,0.26416977246602374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,7168,0.5674542321099175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,8192,0.7126408682929145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,51200,1.771132469177246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,7168,0.6000355614556206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,7168,0.22726665602789986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,6144,0.1989848878648546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,5120,0.1696986622280545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,51200,3.9917352464463978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,51200,5.101241641574435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,3584,0.32711021105448407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,6144,0.5475795533921983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,4096,0.3459484312269423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,4096,0.14027289549509683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,3072,0.2215564383400811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,3584,0.3101502259572347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,4096,0.3512248992919922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,3072,0.11055555608537461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,5120,0.4300977653927273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,3072,0.2576346662309435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,2560,0.2385493384467231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,2048,0.1556897825664944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,2560,0.09633421897888184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,2560,0.21931466791364881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,1536,0.1377511156929864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,2048,0.17885777685377333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,1024,0.0979493326610989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,3584,0.12835555606418186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,1536,0.14169955253601074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,6144,0.5290986696879069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,768,0.06929422087139554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,1024,0.10668622122870551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,512,0.05542666713396708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,768,0.046373334195878774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,768,0.08622755606969197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,256,0.03949866692225138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,2048,0.08232177628411187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,1536,0.0641626649432712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,256,0.03979911075698005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,128,0.03259199857711792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,5120,0.45501867930094403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,1024,0.05111111203829447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,128,0.038522667355007596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,128,0.0633173320028517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,64,0.06319200330310397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,512,0.042993777328067355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,512,0.07495822509129842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,256,0.0681777795155843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,32,0.06347022453943889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,64,0.02788355615403917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,32,0.02880711025661892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,16384,0.96051025390625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,16384,1.1685146755642362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,51200,1.278066635131836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,16384,0.39040088653564453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,65536,1.8198729621039496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,51200,3.1086942884657116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,12288,0.7140524652269152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,12288,0.299890677134196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,65536,4.029961903889974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,10240,0.7010062005784777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,12288,1.025302251180013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,10240,0.24748889605204263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,8192,0.543465773264567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,10240,0.6858124203152127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,51200,3.908411661783854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,7168,0.4694168832567003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,7168,0.18267644776238334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,6144,0.3835582203335232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,65536,5.120360056559245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,8192,0.5433635711669922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,6144,0.18158755037519667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,5120,0.312084436416626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,4096,0.27809421221415204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,5120,0.34380443890889484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,3584,0.23482044537862143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,8192,0.23045423295762804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,3072,0.2293191221025255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,3584,0.2334995534684923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,7168,0.47445154190063477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,6144,0.40192445119222003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,3072,0.08727466397815281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,2560,0.1659653319252862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,5120,0.13471110661824545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,2048,0.12158132923973931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,4096,0.11266667313045925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,4096,0.26965154541863334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,2560,0.17223644256591797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,2048,0.06505955590142144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,2048,0.1417377789815267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,1024,0.08668710788091023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,1536,0.11215910646650527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,1024,0.0406951109568278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,768,0.05979644589953952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,3072,0.20377600193023682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,768,0.03761511047681173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,512,0.058267553647359215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,768,0.06866221957736544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,2560,0.07664533456166585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,512,0.03511200017399258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,512,0.06001066499286228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,128,0.02648177742958069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,256,0.032615999380747475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,3584,0.10313244660695393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,64,0.022589332527584497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,1536,0.05147288905249702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,32,0.023355555203225877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,128,0.05082755618625217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,1024,0.08449955781300862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,64,0.05077244506941902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,256,0.03408266769515143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,1536,0.11394666300879584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,256,0.054530666934119336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,128,0.03187200095918443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,65536,0.5673599772983128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,65536,1.3609502580430772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,51200,1.0329653422037761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,16384,0.39843021498786074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,51200,0.429620451397366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,12288,0.2636648813883464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,16384,0.13470044400956896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,16384,0.3743164539337158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,12288,0.10246577527787949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,10240,0.22303999794854057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,10240,0.08689244588216145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,10240,0.22237687640719941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,8192,0.0723137789302402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,8192,0.17669333351982963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,65536,1.5955822202894423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,7168,0.19295644760131836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,51200,1.2182515462239583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,12288,0.2814142174190945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,6144,0.147507561577691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,7168,0.06309955649905734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,7168,0.1555537780125936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,5120,0.12374222278594971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,6144,0.05690755446751913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,8192,0.18278400103251138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,4096,0.10944088962343003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,5120,0.047467556264665395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,6144,0.1356782250934177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,3584,0.09499644570880467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,4096,0.03949777616394891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,3072,0.07729599873224895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,32,0.05102755626042684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,3584,0.08162666691674127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,3072,0.03179733289612664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,3072,0.07163200113508436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,2048,0.051503111918767296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,2560,0.02827733423974779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,2560,0.06235466400782267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,1536,0.043450666798485645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,2048,0.05076800121201409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,1024,0.02720977697107527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,1536,0.020576889316240948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,4096,0.09187377823723687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,1536,0.04081333345837063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,768,0.021571555071406897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,1024,0.017272889614105225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,512,0.01696266730626424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,768,0.015722667177518208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,768,0.02569777766863505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,256,0.013219555219014486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,512,0.02169955604606205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,2048,0.024547555380397375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,256,0.013412444127930535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,256,0.019301333361201815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,5120,0.11423910988701715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,64,0.009284444153308868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,3584,0.035536001125971474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,32,0.009308444129096137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,1024,0.03118577930662367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,64,0.018290667070282828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,32,0.01832355558872223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,128,0.01057866629627016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,65536,0.9325520197550455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,128,0.013459555804729462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,128,0.018736888964970905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,16384,0.2631982167561849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,65536,1.183526251051161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,512,0.014548444085650973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,51200,0.33872267935011124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,12288,0.21703733338250053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,16384,0.3027057647705078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,12288,0.0831751094924079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,10240,0.17146666844685873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,2560,0.06519289149178399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,8192,0.12982222768995497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,10240,0.17308089468214247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,8192,0.05686489078733656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,16384,0.10478044218487209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,7168,0.1299617820315891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,8192,0.13547200626797146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,51200,0.962831073337131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,7168,0.05082044336530897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,65536,0.4319359991285536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,7168,0.12000889248318142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,6144,0.044215109613206655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,10240,0.06846221950319079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,6144,0.10444266928566827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,51200,0.7484230995178223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,5120,0.03752355443106757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,5120,0.08820799986521403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,4096,0.031199110878838435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,4096,0.07181244426303439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,3072,0.052389333645502724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,3584,0.028573334217071533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,3584,0.06330488787757026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,12288,0.20569777488708496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,6144,0.10149600108464558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,3072,0.025460445218616064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,2048,0.04331199990378486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,5120,0.10622933175828721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,2560,0.04831377665201823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,4096,0.08297777838177152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,2048,0.020250666472646926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,2048,0.04011911153793335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,1024,0.021831111775504217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,1536,0.033245331711239286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,1024,0.014519110321998596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,2560,0.05256355471081204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,768,0.01753777762254079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,3072,0.05620977613661024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,768,0.013292444248994192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,512,0.01389600005414751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,3584,0.07674311267005073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,1536,0.01717155509524875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,256,0.010052444206343757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,512,0.01795733306143019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,1024,0.025009777810838487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,256,0.011731555892361535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,256,0.015584889385435315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,2560,0.023303111394246418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,128,0.011763555308183035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,768,0.02090488870938619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,32,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,64,0.015204444527626038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,512,0.012484444512261285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,32,0.015180443723996481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,128,0.008347555167145198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,65536,0.7765937911139594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,65536,0.3547653357187907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,1536,0.028958221276601154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,51200,0.6445857683817545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,16384,0.23689421017964682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,65536,1.0441617965698242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,12288,0.1718906693988376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,16384,0.08908000257280137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,16384,0.22566488054063585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,51200,0.7855342229207357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,64,0.007618666523032718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,12288,0.06866221957736544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,12288,0.1689173380533854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,10240,0.059676441881391734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,10240,0.14281155003441706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,8192,0.048710223701265126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,8192,0.11664177311791314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,51200,0.2789706654018826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,7168,0.04327377676963806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,7168,0.1016319990158081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,5120,0.08931822246975368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,6144,0.08898044294781154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,10240,0.17359022299448648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,128,0.015192889504962497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,4096,0.07045333253012763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,5120,0.033019555939568415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,3584,0.06210755639606052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,7168,0.10831555393007065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,4096,0.06162399715847439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,3584,0.025177778469191656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,3072,0.04815999997986687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,3584,0.0551128884156545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,2560,0.0421635541650984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,3072,0.022780444886949327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,3072,0.04799555407630073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,2560,0.02068800065252516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,2048,0.03394577900568644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,2560,0.042057779100206166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,8192,0.14443911446465388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,5120,0.07490933603710599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,2048,0.01834222177664439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,1024,0.019655999210145738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,1536,0.015789333317014907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,1536,0.02867377797762553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,1024,0.013584888643688627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,1024,0.021822222405009802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,6144,0.03783555494414435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,768,0.012268444730175866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,768,0.01815733313560486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,512,0.011447110937701331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,1536,0.02608888844648997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,512,0.015503111812803479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,2048,0.034854223330815635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,128,0.007418666448858049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,768,0.015526221858130561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,256,0.013921777407328287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,64,0.006614222294754452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,128,0.010814222196737925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,128,0.01353422221210268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,64,0.013460444079505073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,32,0.013439111411571503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,4096,0.02749866743882497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,256,0.010795555180973478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,6144,0.10284622510274251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,65536,0.719505786895752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,65536,0.30714133050706655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,32,0.00666933341158761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,51200,0.5048640039232042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,16384,0.1885839965608385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,51200,0.23874311976962617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,12288,0.15459288491143122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,16384,0.07539466354582045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,256,0.008926221893893348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,10240,0.11184267203013103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,12288,0.06318755282296075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,8192,0.10060977935791016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,10240,0.05049599872695076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,10240,0.11750488811069065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,512,0.01239466667175293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,8192,0.040433777703179255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,8192,0.09629155529869927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,65536,0.7982293234931098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,7168,0.03577422102292379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,6144,0.07617066966162787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,7168,0.08391288916269939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,51200,0.6283351050482856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,16384,0.18417244487338594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,5120,0.06523822413550483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,6144,0.07351733578575982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,4096,0.05336977707015144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,5120,0.028324445088704426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,5120,0.06258933411704169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,3584,0.046012444628609546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,4096,0.050705777274237744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,7168,0.08251466353734334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,3072,0.0370435549153222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,3584,0.04477777745988634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,2560,0.033294220765431724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,3072,0.020029332902696397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,3072,0.039650665389166936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,2048,0.026597332623269822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,6144,0.03222222129503886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,2560,0.03489689032236735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,1536,0.022942221826977197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,12288,0.14128000206417507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,2048,0.029471112622155085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,1536,0.013783111340469785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,1536,0.024075556132528517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,1024,0.012080888781282636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,768,0.0148062225845125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,2560,0.018401778406567044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,512,0.012081777883900536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,768,0.015676443775494892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,2048,0.016137777103318107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,256,0.008604444563388824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,512,0.010394666757848527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,512,0.01346666696998808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,128,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,3584,0.022110222114457026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,256,0.010042666561073726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,64,0.006562666760550604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,128,0.010020444790522257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,128,0.01182755579551061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,64,0.011740444435013665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,32,0.011763555308183035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,4096,0.023935111032591924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,1024,0.017151999804708693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,65536,0.6112435658772787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,256,0.012265777422322167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,1024,0.01856355534659492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,51200,0.454635567135281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,16384,0.16170667277442083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,65536,0.7072035471598307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,51200,0.19881867037879097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,12288,0.1019573344124688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,768,0.011301333705584208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,51200,0.5957377751668295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,10240,0.10143822431564331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,12288,0.05080266793568929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,12288,0.12456356154547797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,8192,0.07629333602057563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,10240,0.04283110962973701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,10240,0.10508977704577976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,7168,0.0712862213452657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,65536,0.2603742281595866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,8192,0.08528266350428264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,6144,0.06529155704710218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,7168,0.07521688938140869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,6144,0.027951111396153767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,5120,0.05167999863624573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,16384,0.06342133548524645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,16384,0.18145955933464897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,4096,0.04114133450720046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,5120,0.024654222859276667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,5120,0.05558400021659004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,4096,0.0214675549003813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,4096,0.045003556542926364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,3584,0.01966399947802226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,8192,0.03451111250453525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,3584,0.04031555520163642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,7168,0.031031999323103163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,3072,0.01802399920092689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,3072,0.03660000032848782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,2048,0.021681777305073206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,6144,0.06548533174726698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,2560,0.031480888525644936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,1536,0.016993777619467843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,2048,0.014380445082982382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,3584,0.03710133168432448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,2048,0.027087999714745417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,1024,0.013000888956917657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,1536,0.012776888906955719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,1536,0.022164444128672283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,3072,0.03092977735731337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,32,0.006703111032644908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,1024,0.016777777009540133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,2560,0.028024888700909082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,768,0.010423999693658618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,768,0.014503111441930136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,512,0.009734222458468543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,2560,0.016407110624843173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,512,0.012434666355450949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,128,0.006320888797442119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,256,0.011405333048767514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,128,0.009343110852771336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,128,0.011071110765139261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,64,0.005585777676767773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,32,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,512,0.008707555631796518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,64,0.011032889286677042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,256,0.006923555499977536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,32,0.011075555450386472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,256,0.009347555538018545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,65536,0.45338577694363064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,65536,0.24116534656948516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,51200,0.4441368844774034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,16384,0.1384506622950236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,65536,0.6135359870062934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,51200,0.18771555688646105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,768,0.010517333116796283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,12288,0.10123644272486369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,16384,0.060897780789269336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,51200,0.47910488976372617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,10240,0.09736355808046128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,12288,0.04760533240106371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,8192,0.0770604411760966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,10240,0.043543110291163124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,12288,0.11236266295115153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,10240,0.09290933609008789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,8192,0.03380711211098565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,8192,0.07581866449779935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,7168,0.030044443077511255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,7168,0.06719288561079237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,6144,0.027154665854242113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,6144,0.0583217806286282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,1024,0.011110222174061669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,4096,0.0395342210928599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,5120,0.024067555864651997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,16384,0.20580266581641304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,4096,0.020735111501481798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,4096,0.04100355505943298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,3584,0.01889866590499878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,3584,0.036935110886891685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,6144,0.056582219070858426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,2560,0.023923556009928387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,3072,0.03311466508441501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,5120,0.04888800117704603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,2048,0.020727111233605277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,2560,0.02936977810329861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,5120,0.04960888955328199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,1536,0.016695110334290397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,2048,0.024693333440356787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,3072,0.029190222422281902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,1024,0.012429333395428128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,1536,0.012490666574902005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,1536,0.02033244404527876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,768,0.010612444745169746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,3072,0.017318222257826064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,1024,0.010833777487277985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,512,0.0083155557513237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,768,0.010030222435792288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,2560,0.015746666325463187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,256,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,3584,0.035255110926098294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,512,0.011752000285519494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,2048,0.014147554834683737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,256,0.009114666945404476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,7168,0.06804000006781684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,64,0.0052373334765434265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,128,0.00903555585278405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,128,0.010230222509966956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,1024,0.015338665909237333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,32,0.010049777726332346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,768,0.013112000293201871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,512,0.009582222335868413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,128,0.005633777628342311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,65536,0.21999733977847627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,256,0.010388444695207808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,65536,0.5197893248664008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,32,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,51200,0.1752880016962687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,64,0.010091555615266165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,12288,0.09370666742324829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,16384,0.12623022662268743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,12288,0.04430933462248909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,12288,0.09473244349161784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,65536,0.47640710406833225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,8192,0.06239555279413859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,10240,0.03768622212939792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,10240,0.08119733466042413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,16384,0.11771821975708008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,8192,0.030243555704752605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,8192,0.0662684440612793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,51200,0.40088356865776914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,6144,0.04667377803060743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,7168,0.05894044372770521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,5120,0.04096800088882446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,10240,0.07415999968846639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,6144,0.05113066567314995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,4096,0.03282399972279867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,5120,0.04386577672428555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,4096,0.03641422258483039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,7168,0.052541332112418286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,16384,0.055832001898023814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,4096,0.01924000018172794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,7168,0.02716000046994951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,3072,0.026058667235904272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,3584,0.03265155686272515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,5120,0.021899556120236714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,3072,0.015801777442296345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,3072,0.029438220792346533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,2560,0.014083556003040738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,2048,0.018608889646000333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,2560,0.025880000657505457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,3584,0.028408000866572063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,1536,0.014816888504558139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,2048,0.013208000196350945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,2048,0.02195555633968777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,1024,0.011037333144081963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,1536,0.017661333084106445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,2560,0.021577777134047613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,51200,0.38545865482754177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,768,0.009679111341635386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,768,0.00906488878859414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,512,0.0075297777851422625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,768,0.012106666962305704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,256,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,6144,0.024890666206677754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,512,0.0107351111041175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,1536,0.011100444528791638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,128,0.005587555468082428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,256,0.008625778059164682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,256,0.009704888694816166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,128,0.00832444429397583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,32,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,128,0.009102221992280748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,64,0.009030222064918941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,32,0.009387555221716562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,65536,0.38120177057054305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,512,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,3584,0.016943999462657504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,65536,0.17446666293674043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,65536,0.43210577964782715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,64,0.005081777771313985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,51200,0.13792444599999323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,1024,0.009718221922715506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,51200,0.33320800463358563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,16384,0.04725955592261421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,16384,0.10559733046425714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,12288,0.037124445041020714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,12288,0.08020799689822726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,10240,0.03299555513593886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,8192,0.051671998368369214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,10240,0.0684444440735711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,51200,0.27679112222459584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,7168,0.04215555720859104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,8192,0.05542400148179796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,6144,0.0384168889787462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,7168,0.0244862222009235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,1024,0.013797333670987023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,7168,0.04954399996333652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,5120,0.032885332902272545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,6144,0.0222053329149882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,10240,0.06519999768998888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,6144,0.04333422250217862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,5120,0.020041777027977836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,3584,0.024075556132528517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,4096,0.017087111870447796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,4096,0.03123733401298523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,3584,0.01570933394961887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,8192,0.026791110634803772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,3584,0.028247111373477515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,2560,0.018944000204404194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,3072,0.014208000567224292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,3072,0.025264889001846313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,12288,0.07415199942058988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,2560,0.013247999880048962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,2560,0.02257599929968516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,4096,0.026895999908447266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,5120,0.03730400072203742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,1536,0.015503111812803479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,1024,0.010065777434243096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,16384,0.09722399711608887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,1024,0.009552888572216034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,1024,0.012145777543385824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,1536,0.013212444053755866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,768,0.010826666322019366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,2048,0.01891911029815674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,512,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,1536,0.010738666686746808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,512,0.008448888858159384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,256,0.005663110978073544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,512,0.009506666825877296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,128,0.004953777624501122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,256,0.00814400033818351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,256,0.0086666668454806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,768,0.00830488900343577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,128,0.008067555725574493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,32,0.004722666823201709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,128,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,2048,0.012106666962305704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,64,0.008341333104504479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,768,0.009001777403884465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,32,0.008359111017651027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,3072,0.021172444025675457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,65536,0.14411644140879312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,65536,0.3813760015699599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,2048,0.015791111522250705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,16384,0.08503555589252049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,51200,0.3038160006205241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,12288,0.06219377782609728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,16384,0.03758577836884393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,16384,0.0946719977590773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,12288,0.03050400151146783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,12288,0.07276978095372517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,51200,0.2682737774319119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,10240,0.02604444490538703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,8192,0.043046222792731396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,10240,0.061892443233066134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,7168,0.03825955589612325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,51200,0.11477066410912408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,8192,0.021347555849287245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,6144,0.033851553996404014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,7168,0.019934222102165222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,7168,0.04541688826349047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,6144,0.0181697772608863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,65536,0.2647884421878391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,6144,0.0391813317934672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,4096,0.02163022259871165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,5120,0.016307555966907077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,3584,0.020185778538386028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,4096,0.028580443726645574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,64,0.00461777796347936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,3072,0.017294221454196505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,8192,0.050473776128556996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,3584,0.026120889517996047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,2560,0.015484443969196744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,3072,0.012102222277058495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,5120,0.02688177757793003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,10240,0.054782221714655556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,2048,0.013256000147925483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,2560,0.020469332734743755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,1536,0.011101333631409539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,4096,0.014158222410413953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,2048,0.01719911065366533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,1024,0.008255110846625434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,1536,0.009368889033794403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,1536,0.014452444182501899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,3584,0.01313688854376475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,1024,0.008359111017651027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,1024,0.011400889191362592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,512,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,768,0.008369777765538957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,768,0.010027555955780877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,2560,0.011077333655622272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,256,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,512,0.00904444439543618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,5120,0.033979554971059166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,128,0.004562666846646203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,256,0.007616889145639207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,64,0.004228444563017951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,256,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,128,0.007583111524581909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,128,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,768,0.007010666860474481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,64,0.00793599999613232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,32,0.008001777860853408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,65536,0.30774222479926217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,3072,0.023228444986873206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,65536,0.13580889172024196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,512,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,65536,0.33505866262647843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,51200,0.10575466685824925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,12288,0.05340711275736491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,16384,0.03635644581582811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,32,0.0042257776690853965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,16384,0.08465777503119574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,10240,0.04753066764937507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,12288,0.029162665208180744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,12288,0.06480799780951606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,10240,0.02599999970859951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,10240,0.05488711264398363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,7168,0.032297776805029974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,51200,0.2332951227823893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,8192,0.04530310961935255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,16384,0.07395288679334852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,7168,0.01942399972014957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,7168,0.04055111275778876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,51200,0.27225155300564235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,5120,0.023471110396915015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,6144,0.03554933269818624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,4096,0.019350222415394254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,5120,0.030797332525253296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,4096,0.01384622189733717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,4096,0.0259226659933726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,3584,0.018018666240904067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,6144,0.028255999088287354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,3584,0.013152889079517789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,3072,0.015821332732836407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,3584,0.023324444890022278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,2560,0.013403555585278405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,6144,0.017900443739361234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,3072,0.020962667134073045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,5120,0.015885333220163982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,2048,0.011750222080283694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,2560,0.011003555523024665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,2560,0.01794399983353085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,1536,0.009685333404276106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,2048,0.009879111415810054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,8192,0.021332444416152105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,2048,0.015631111131774057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,1536,0.009319999979601966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,1536,0.013171555267439948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,2048,0.010023111270533668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,768,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,1024,0.008546666966544257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,512,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,3072,0.011965333587593503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,768,0.008016000191370646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,256,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,768,0.009319999979601966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,512,0.008000888758235509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,128,0.004212444441186057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,512,0.00868266655339135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,256,0.007489778101444244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,64,0.004039111236731212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,128,0.007374222079912822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,32,0.00424888895617591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,128,0.007420444654093848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,64,0.00740444411834081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,65536,0.20295733875698516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,32,0.0074755557709270054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,1024,0.010449777874681683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,51200,0.20570577515496147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,65536,0.13367644945780435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,16384,0.06399199697706434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,51200,0.11064089006847805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,51200,0.23136978679233125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,12288,0.045780443482928805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,16384,0.03660710983806186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,1024,0.007637333538797166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,10240,0.04233333468437195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,12288,0.029071112473805744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,12288,0.056709335909949414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,8192,0.032848887973361544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,10240,0.04832622077729967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,7168,0.028522666957643297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,8192,0.0221404449807273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,8192,0.04006400042110019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,256,0.007672888537247975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,7168,0.01923733287387424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,7168,0.036032888624403216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,5120,0.021723555194007024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,6144,0.03166844447453817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,10240,0.02571822206179301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,4096,0.018627555833922494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,5120,0.015513777732849121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,65536,0.2930240101284451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,3584,0.016753777861595154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,6144,0.025376889440748427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,16384,0.07447733481725057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,3584,0.012979555461141797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,6144,0.017431111799346078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,3072,0.011824888487656912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,2560,0.013377777404255338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,3072,0.018383110562960308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,5120,0.02754488918516371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,2048,0.011408888631396823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,2560,0.01073066641887029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,2560,0.016389333539538913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,1536,0.009708444277445475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,4096,0.022983110613293115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,2048,0.014171555638313293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,1024,0.0074142225914531285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,3072,0.014904888139830695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,1536,0.012005333271291522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,768,0.0069520001610120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,1024,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,1024,0.009757333331637913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,512,0.005667555663320754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,768,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,768,0.00868977771864997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,8192,0.0377439988984002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,512,0.007648888561460707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,4096,0.013442666994200813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,2048,0.009719111025333405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,512,0.00811111099190182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,1536,0.009031111167536842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,3584,0.02073955535888672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,64,0.003928889003064898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,256,0.007294221884674496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,32,0.0038844446341196695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,64,0.007247111035717859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,65536,0.22854932149251303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,32,0.007132444116804335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,256,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,51200,0.1557119952307807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,128,0.004217777815130022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,256,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,65536,0.2471484343210856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,51200,0.10023733642366196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,51200,0.1966951158311632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,128,0.007458666960398356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,16384,0.06373688909742567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,12288,0.02760800056987339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,10240,0.035312000248167254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,65536,0.12753689289093018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,10240,0.041793776883019336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,10240,0.024824000067181055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,16384,0.03400977783732944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,128,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,12288,0.04273866613705953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,7168,0.025238222546047632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,8192,0.021348444951905146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,6144,0.022095999783939783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,7168,0.03144888745413886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,6144,0.016569778323173523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,6144,0.027943111128277246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,5120,0.01908533275127411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,16384,0.055071999629338585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,8192,0.02865955564710829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,5120,0.014715555641386243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,5120,0.024148444334665935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,3584,0.01548266741964552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,8192,0.034661332766215004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,7168,0.018351111147138808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,3584,0.012409778104888068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,3584,0.01812533371978336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,2560,0.01144088887506061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,12288,0.04898222287495931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,3072,0.016360888878504436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,2048,0.010374222364690568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,4096,0.015824000040690105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,2560,0.010739555789364709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,1536,0.008820444345474243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,4096,0.013295110729005603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,4096,0.020084444019529555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,1024,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,1536,0.008791111409664154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,1536,0.010808000134097205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,768,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,1024,0.008065777520338694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,1024,0.008760888543393876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,3072,0.012095111111799875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,768,0.007668444679843054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,768,0.008023111356629265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,256,0.004228444563017951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,512,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,512,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,2560,0.014447111222479077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,128,0.003907555507289039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,256,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,64,0.0035546666218174826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,128,0.006984000404675801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,32,0.003547555456558863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,3072,0.013433777623706393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,128,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,64,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,65536,0.1725262271033393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,512,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,51200,0.1342328919304742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,65536,0.2041377756330702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,16384,0.044122666120529175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,51200,0.16291466024186876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,256,0.006776000062624614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,16384,0.023392889234754775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,12288,0.03580088747872247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,16384,0.05346044566896227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,10240,0.028099556763966877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,12288,0.020063999626371596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,65536,0.08038489023844402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,8192,0.023370666636361018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,10240,0.018139556050300598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,2048,0.009713778065310584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,10240,0.035306665632459856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,2048,0.01258399999803967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,8192,0.029483556747436523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,8192,0.01458755632241567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,7168,0.013460444079505073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,6144,0.012513777448071374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,5120,0.016487111647923786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,32,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,6144,0.023676445086797077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,4096,0.013803555733627744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,5120,0.011408888631396823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,5120,0.020238222347365487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,3584,0.012449777788586087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,51200,0.06492977672153048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,7168,0.02115999990039402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,4096,0.016993777619467843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,3584,0.009720000127951304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,6144,0.01867466668287913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,3584,0.015563555889659457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,3072,0.009390222529570261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,3072,0.014182221558358936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,7168,0.02695911129315694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,12288,0.04103199972046746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,2048,0.008325333396593729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,1536,0.00721244431204266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,4096,0.010344888601038191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,3072,0.01107022249036365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,1536,0.007991111112965478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,1536,0.009679111341635386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,1024,0.005972444597217772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,1024,0.0073724447025193115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,2048,0.008464888566070134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,2560,0.012768000364303589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,768,0.007068444457319047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,2048,0.01110666659143236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,2560,0.010036444498433007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,512,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,2560,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,512,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,256,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,128,0.00360000009338061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,512,0.006766222417354584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,128,0.006676444576846228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,128,0.005987555616431766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,32,0.003548444559176763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,64,0.006167111297448476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,768,0.007343110938866933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,32,0.006064888917737537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,65536,0.14703555901845297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,51200,0.11580888430277507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,65536,0.07929599947399564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,65536,0.1570728884802924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,16384,0.04203022188610501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,51200,0.06405511167314318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,1024,0.008001777860853408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,51200,0.1244631078508165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,12288,0.0311155551009708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,16384,0.022480888499153986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,10240,0.02702666653527154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,16384,0.042075554529825844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,12288,0.018752000398106046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,256,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,8192,0.02307555576165517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,10240,0.017472000585661996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,7168,0.018008889423476327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,8192,0.014136888914638095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,8192,0.023930667175187006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,6144,0.018583999739752877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,7168,0.021683555510309007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,768,0.005562666803598404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,5120,0.014332445131407844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,6144,0.012335999972290464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,5120,0.011426666544543372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,4096,0.012103111379676394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,5120,0.016593777471118502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,10240,0.028453333510292902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,4096,0.010055999788973067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,4096,0.014204444156752693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,7168,0.01328888866636488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,3072,0.00979288915793101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,3584,0.009681777821646797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,3584,0.012958221965365939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,64,0.003579555700222651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,2560,0.00868711123863856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,6144,0.019098665979173448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,3072,0.009163555171754625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,2048,0.007972444097201029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,2560,0.008661333057615492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,1536,0.007305777735180325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,12288,0.032949331733915545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,2560,0.010850666297806634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,1536,0.007873777714040544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,3584,0.012415111064910889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,1024,0.007335111498832703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,768,0.004909333255555895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,256,0.0042453333735466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,768,0.006981333096822103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,512,0.004563555535342959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,2048,0.008366222182909647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,512,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,512,0.006161777923504512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,2048,0.009720000127951304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,256,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,1024,0.007325333025720384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,3072,0.011815110842386881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,768,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,64,0.003195555466744635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,128,0.006610666712125142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,32,0.0032035555276605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,1024,0.005680000202523337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,1536,0.00865955568022198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,64,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,65536,0.09609333011839126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,32,0.00564444437623024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,51200,0.07738488912582397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,65536,0.0789546635415819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,256,0.005787555542257097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,16384,0.03035822179582384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,51200,0.06333421998553805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,51200,0.1222462256749471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,16384,0.022081777453422546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,128,0.005576888720194499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,16384,0.041061331828435264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,10240,0.0243484444088406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,12288,0.03250311149491204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,256,0.0036595554815398324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,10240,0.017156443662113614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,128,0.0032213332338465583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,8192,0.023015111684799194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,65536,0.15273243851131862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,8192,0.014019555515713163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,7168,0.012969777815871768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,6144,0.01586933268441094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,12288,0.01847999956872728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,8192,0.02018933329317305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,5120,0.013493333425786761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,6144,0.012103111379676394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,6144,0.018801778554916382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,4096,0.010176888770527309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,5120,0.011064888702498542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,5120,0.016523554921150208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,3584,0.009391999906963771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,4096,0.010060444474220276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,4096,0.013996443814701505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,3072,0.008584000170230865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,3584,0.012936000194814471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,10240,0.028127110666698877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,3072,0.009015999734401703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,2560,0.01072622256146537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,2048,0.006960888703664144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,12288,0.027816888358857896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,3584,0.009555555880069733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,2048,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,1536,0.006000888844331105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,3072,0.011745777395036487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,1024,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,7168,0.016639999217457242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,2560,0.00867022242810991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,1024,0.00699733363257514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,1024,0.007351111206743453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,768,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,512,0.003938666648334927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,2560,0.008010666403505538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,1536,0.008385777473449707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,512,0.006688888702127669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,768,0.004569777597983678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,256,0.005621333503060871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,128,0.0032551110618644287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,2048,0.009712888962692684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,7168,0.020962667134073045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,64,0.003198222153716617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,128,0.005502222312821283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,32,0.0032026666320032547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,512,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,64,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,65536,0.07146400213241577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,32,0.005791999813583162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,65536,0.07931022511588202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,65536,0.11112266116672093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,256,0.006490666833188798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,16384,0.023112888137499493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,51200,0.06351200077268812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,128,0.006517333288987477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,12288,0.019165333774354722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,16384,0.021609778205553692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,16384,0.030952887402640447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,256,0.0035991109907627106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,12288,0.018251554833518136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,12288,0.024713777833514746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,8192,0.01366666704416275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,10240,0.016587555408477783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,51200,0.05795200003517998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,10240,0.02141155633661482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,8192,0.01351466692156262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,8192,0.018165333403481376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,768,0.006643555644485686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,7168,0.012764444781674279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,1536,0.007672888537247975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,5120,0.011374221907721626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,6144,0.014936000108718872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,5120,0.010783111055692038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,4096,0.008358221915033128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,7168,0.012808000048001608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,4096,0.009689778089523315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,4096,0.01167555567291048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,7168,0.016730666160583496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,6144,0.011757333245542316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,3072,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,10240,0.016686222619480558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,3584,0.009530666801664565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,3584,0.010775999890433418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,2560,0.006948444578382704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,5120,0.013459555804729462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,2048,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,2560,0.008592888712882996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,6144,0.011058666639857821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,2048,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,3584,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,2048,0.008059555457697975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,1024,0.004586666822433472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,1536,0.007633777956167857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,1536,0.007343110938866933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,768,0.0038986665507157645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,1024,0.0063075555695427795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,3072,0.00905600024594201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,512,0.0038986665507157645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,768,0.0069573331210348345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,768,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,512,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,2560,0.009184888667530483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,1536,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,256,0.006324444380071428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,128,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,1024,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,256,0.00535733335547977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,128,0.0063279999627007385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,128,0.005246222019195557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,32,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,64,0.005209777918126848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,512,0.005648889061477449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,51200,0.08753955364227295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,65536,0.07891111241446601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,65536,0.11099200116263495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,51200,0.03479822145568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,64,0.002959111084540685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,3072,0.010045333041085137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,16384,0.01569955547650655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,12288,0.013806222213639153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,16384,0.02119111186928219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,256,0.0031893334041039148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,10240,0.012136889000733694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,12288,0.0181351105372111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,12288,0.02403733299838172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,8192,0.011590222517649332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,10240,0.01661688917213016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,51200,0.0629297759797838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,7168,0.011519110865063138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,8192,0.01365244471364551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,8192,0.017904000149832833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,32,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,6144,0.010478222535716163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,6144,0.011800000237094032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,5120,0.009097778134875828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,10240,0.020738666256268818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,5120,0.010999110837777456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,51200,0.08725688854853313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,4096,0.007985778152942657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,4096,0.009707555174827576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,16384,0.030231111579471167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,65536,0.04235555397139656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,7168,0.012708444562223224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,3072,0.006656000183688269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,3584,0.009342222577995723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,3584,0.010565333068370819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,6144,0.014764444695578681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,2560,0.006011555592219035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,3072,0.009008000294367472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,2048,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,2560,0.008452444440788692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,5120,0.013125333521101209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,1536,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,2048,0.008052444292439355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,3584,0.007288888924651676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,7168,0.016200888488027785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,1024,0.0042151109211974675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,1536,0.007338666253619724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,768,0.0038542221817705366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,1024,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,1024,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,768,0.006940444310506185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,3072,0.009734222458468543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,768,0.006067555397748947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,2560,0.009015999734401703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,512,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,256,0.006317333214812809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,256,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,128,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,512,0.003564444267087512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,64,0.0028728888266616394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,128,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,32,0.0029440000653266907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,256,0.003334222154484855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,512,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,65536,0.032969776127073504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,64,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,51200,0.027319111757808264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,32,0.004965333475006951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,4096,0.011414222419261932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,16384,0.012591111163298288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,65536,0.10881955756081475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,51200,0.06263911061816745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,51200,0.08549066384633382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,16384,0.021120000216695998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,16384,0.029467556211683486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,10240,0.010032888915803697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,12288,0.02349066734313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,2048,0.008008889026112026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,10240,0.01628888812330034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,1536,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,8192,0.01756355497572157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,7168,0.008660444782839881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,12288,0.018062222335073683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,7168,0.016141333513789706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,7168,0.012616000241703458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,8192,0.009024000002278222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,10240,0.020566221740510728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,5120,0.00833422193924586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,65536,0.07843733496136136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,6144,0.014656000667148165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,8192,0.013456888496875763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,5120,0.010729778144094678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,12288,0.0107342220014996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,5120,0.013034666577974955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,3584,0.007292444507280986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,4096,0.009684444301658207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,4096,0.011344888971911537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,3584,0.009341333475377824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,6144,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,3584,0.010442666709423065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,2560,0.005979555762476391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,3072,0.009018667042255402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,128,0.0029075555503368378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,3072,0.009680888719028896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,2048,0.005375999957323074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,4096,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,1536,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,2048,0.008232889076073965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,1536,0.007632888853549957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,3072,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,1024,0.004233777936961916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,1536,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,768,0.0035982223020659555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,1024,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,512,0.0035502223504914176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,768,0.006955555743641323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,2560,0.008362666600280339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,768,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,256,0.0032177778581778207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,2560,0.009137777818573846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,2048,0.008051555189821456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,128,0.002983111060327954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,256,0.005290666802061929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,64,0.0027031112048361036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,128,0.006411555740568373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,32,0.0031759999692440033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,128,0.005017777697907554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,64,0.004933333231343163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,65536,0.03125866585307651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,6144,0.011748444702890186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,1024,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,51200,0.024715556038750544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,16384,0.011961778004964193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,12288,0.010328888893127441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,32,0.005045333256324132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,65536,0.10982044537862141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,10240,0.009687110781669617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,256,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,8192,0.008789333204428354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,12288,0.023406222462654114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,10240,0.020270221763186984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,8192,0.017332444588343304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,7168,0.016190222567982145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,5120,0.008665777742862701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,6144,0.014494222071435718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,4096,0.007594666547245449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,5120,0.01292622254954444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,51200,0.0856248868836297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,3584,0.007287110719415877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,3072,0.006598222172922558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,7168,0.008446222378147973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,2560,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,6144,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,2048,0.005233777893914117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,3072,0.009676444033781687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,1536,0.004573333180612988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,2560,0.008933333059151968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,1024,0.003916444463862313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,2048,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,768,0.00388711111413108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,1536,0.007296889192528195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,512,0.005595555735958948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,1024,0.006320000108745363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,4096,0.011186666786670685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,16384,0.029391109943389893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,128,0.0028897778441508612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,3584,0.010391111175219217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,64,0.0025111111915773815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,128,0.004958222309748332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,32,0.002586666701568498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,512,0.006614222294754452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,65536,0.026142223013771906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,512,0.0035457776652442086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,32,0.004976888911591636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,256,0.0031973332580592898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,768,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,16384,0.013895110951529609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,65536,0.10841688844892715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,12288,0.011887999872366587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,10240,0.01091288857989841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,256,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,16384,0.029042667812771265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,8192,0.010049777726332346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,12288,0.023415999280081853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,7168,0.013456888496875763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,6144,0.012117333710193634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,64,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,5120,0.010059555371602377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,7168,0.015968888998031616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,51200,0.020052444603708055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,6144,0.014471999473041959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,4096,0.008723555339707268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,3584,0.008025777836640676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,512,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,5120,0.012784889174832238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,3072,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,4096,0.011148444480366178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,2560,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,2048,0.005576888720194499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,3072,0.009714666340086196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,1536,0.00489688871635331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,10240,0.020406223005718656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,8192,0.017315554949972365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,1024,0.003930666794379552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,768,0.003581333491537306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,1536,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,512,0.0032604444358083936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,1024,0.0062977779242727495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,256,0.0032177778581778207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,768,0.006060444646411472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,128,0.0028808888875775864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,512,0.005671111245950063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,3584,0.010419555836253695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,64,0.002526222210791376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,256,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,32,0.002625777696569761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,2560,0.008815111385451423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,128,0.005015111217896144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,2048,0.008043555749787224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,32,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,51200,0.08586666981379192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,64,0.00490399988161193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,16384,1.3277307086520724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,16384,0.4210853311750624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,16384,1.3915386199951172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,12288,0.940000851949056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,51200,1.3972773022121852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,12288,0.3200764391157362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,12288,1.0307306713528104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,10240,0.9981813430786133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,16384,1.1792809168497722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,10240,0.3147955470614963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,12288,0.8425111240810819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,51200,4.7190123664008246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,51200,3.8297030131022134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,8192,0.2289431095123291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,7168,0.5043573379516602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,8192,0.6597217983669704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,10240,0.8469733132256402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,8192,0.6158133082919651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,7168,0.5661190880669488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,6144,0.5184275309244791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,10240,0.713814205593533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,6144,0.48496177461412215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,6144,0.17147999339633516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,7168,0.5022524197896322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,7168,0.20417600207858613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,8192,0.5669262144300673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,6144,0.44825156529744464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,5120,0.14811110496520996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,4096,0.37235199080573184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,5120,0.3658275604248047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,3584,0.30810843573676217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,4096,0.32052710321214467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,4096,0.12363733185662164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,5120,0.3573875427246094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,3072,0.22395732667711046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,3584,0.10930844148000081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,5120,0.40151556332906085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,3584,0.27155288060506183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,3072,0.09688444270028009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,2560,0.18798577785491943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,3072,0.2437368763817681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,3072,0.22907466358608672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,2048,0.14832177427079943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,51200,4.0590871175130205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,2560,0.2092426617940267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,1536,0.14667556020948622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,4096,0.3068266709645589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,3584,0.2835582097371419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,2048,0.0721893310546875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,2048,0.16220888826582167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,1024,0.08563466866811116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,1536,0.05749155415429009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,1536,0.13526222440931532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,1536,0.13033333089616564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,1024,0.04611644479963514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,1024,0.10195644696553548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,1024,0.09863377941979302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,768,0.04169777697987027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,2560,0.1950168874528673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,2048,0.17070755693647596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,512,0.07166044579611884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,512,0.03776355584462484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,768,0.06917155451244779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,256,0.06539733542336358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,2560,0.08768444591098362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,128,0.03206844462288751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,256,0.052708443668153554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,768,0.08034311400519477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,128,0.060311112138960093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,256,0.03540088733037313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,128,0.04275288846757677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,128,0.03418755531311035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,32,0.02720977697107527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,64,0.06001777781380547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,512,0.049957331683900624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,256,0.03461155626508925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,32,0.060180445512135826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,64,0.026681777503755357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,512,0.06838755475150214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,768,0.08326133092244466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,65536,1.4334853490193684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,51200,3.233224021063911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,65536,4.1039377848307295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,51200,1.0522080527411568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,16384,1.1216826968722873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,51200,3.608964284261068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,65536,4.7657084994845915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,16384,0.3331377771165636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,12288,0.8691040145026313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,12288,0.7670835389031304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,16384,1.0717688666449654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,10240,0.5860115687052408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,16384,0.926643583509657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,10240,0.6360017988416884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,12288,0.6865279939439562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,12288,0.25770666864183217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,8192,0.5092470910814073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,10240,0.5618266529507107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,8192,0.18046311537424722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,8192,0.5026062329610189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,7168,0.45067999098036027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,7168,0.1564799944559733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,7168,0.46889331605699325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,8192,0.4516257709927029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,7168,0.4063342147403293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,6144,0.13693422741360134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,6144,0.3779111173417833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,51200,3.1682044135199656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,5120,0.33336978488498265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,6144,0.3449022240108914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,5120,0.12404621971978082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,5120,0.3159511089324951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,6144,0.38376355171203613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,65536,3.9938460456000437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,4096,0.09844355450736152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,4096,0.25874310069613987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,5120,0.2883262104458279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,4096,0.24142577913072374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,3072,0.21307288275824654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,3584,0.08703733152813381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,3584,0.2542053328620063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,3072,0.07755377557542589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,4096,0.2487555609809028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,3584,0.20904800626966688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,3584,0.20537422762976754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,2560,0.1643422179751926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,2560,0.15802755620744494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,2048,0.12796799341837564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,10240,0.22164800431993273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,2560,0.173927108446757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,1536,0.1125662194357978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,2048,0.05743733379575941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,2560,0.06810844606823392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,1024,0.06813689072926839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,1536,0.1071831120385064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,1024,0.08141866657469007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,1024,0.07849866814083524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,3072,0.1929182211558024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,768,0.05853866868548923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,1536,0.04576177729500664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,768,0.06618933545218574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,1536,0.10402933094236587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,768,0.034281777011023626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,2048,0.1353208886252509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,2048,0.13066311677296957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,3072,0.18136089377933076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,256,0.05238666799333361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,512,0.04422933194372389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,256,0.028762665059831407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,256,0.04355644517474704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,512,0.054021331999037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,1024,0.03775377737151252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,128,0.04836977852715386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,128,0.02828622195455763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,64,0.02163822286658817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,32,0.02271288964483473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,768,0.06562311119503446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,64,0.04829688866933187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,32,0.04831199844678243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,512,0.057087997595469155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,512,0.031035555733574763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,128,0.03555999861823188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,65536,1.290056864420573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,65536,0.47386757532755536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,51200,0.9753608703613281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,65536,1.441800011528863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,256,0.03086844417783949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,16384,0.3486124409569635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,51200,0.37624356481764054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,51200,1.1910764906141493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,16384,0.11811821990542942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,12288,0.252561781141493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,128,0.024432000186708238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,16384,0.29303733507792157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,12288,0.249034669664171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,12288,0.09091911051008437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,10240,0.21036977238125273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,12288,0.22144622272915312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,10240,0.07590666744444105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,51200,0.9136471218532987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,8192,0.17893244160546196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,16384,0.3349982102711995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,10240,0.19074755244784883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,7168,0.13618844085269505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,8192,0.1671146684222751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,8192,0.1536160045199924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,7168,0.14734489387936062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,6144,0.11829155021243626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,7168,0.13598399692111546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,10240,0.20678222179412842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,6144,0.04932444294293722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,5120,0.12054755952623154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,6144,0.11570666895972358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,8192,0.06118400229348076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,4096,0.07706666655010648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,5120,0.04280444317393833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,5120,0.10691377851698135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,7168,0.05513422356711494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,4096,0.08729777733484904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,4096,0.08232088883717854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,3584,0.07710222403208415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,6144,0.12797955671946207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,3584,0.030840890275107488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,3584,0.07382844554053412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,3072,0.027701333165168762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,3072,0.06809244553248088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,2560,0.06160800324545967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,3072,0.06571911440955268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,5120,0.09915733337402344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,65536,1.2465200424194336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,2560,0.025913778278562758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,2560,0.05849155452516344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,2560,0.057320886188083224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,1536,0.03680355681313409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,2048,0.04900977677769131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,3072,0.07485066519843207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,1536,0.01867111192809211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,1536,0.03911911116706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,1536,0.04106311003367106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,3584,0.08438755406273736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,2048,0.04547199938032362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,768,0.01961599952644772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,1024,0.03277600142690871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,2048,0.02180800007449256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,2048,0.048384000857671104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,768,0.024643555283546448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,768,0.014132445057233175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,768,0.027608889672491286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,1024,0.024361777636739943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,256,0.01274400038851632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,512,0.020983111527231004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,4096,0.035700443718168474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,1024,0.02978133492999607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,128,0.010694222317801582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,1024,0.015623110863897534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,256,0.019505777292781405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,128,0.017880888448821176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,128,0.012127111355463663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,128,0.017249777913093567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,32,0.009319999979601966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,64,0.01795288920402527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,512,0.023057777020666335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,32,0.017921777235137094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,256,0.012454222473833295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,256,0.018572444717089336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,512,0.015219555960761176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,65536,0.9409297307332357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,65536,0.3715537654028998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,51200,0.7338862419128418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,512,0.013038222160604266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,65536,1.1078017552693684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,51200,0.29297243224249947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,16384,0.25123288896348744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,51200,0.8747395409478081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,16384,0.0933519999186198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,12288,0.18778933419121635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,16384,0.24312978320651582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,10240,0.15117243925730386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,12288,0.07021244366963704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,12288,0.1872791051864624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,16384,0.26033687591552734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,8192,0.13081244627634683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,51200,0.755151112874349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,10240,0.06218577755822075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,10240,0.15494933393266466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,8192,0.128729780515035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,7168,0.11840621630350749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,8192,0.12981332672966853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,12288,0.19018310970730254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,7168,0.11276800102657741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,7168,0.04382755690150791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,6144,0.10193866491317749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,6144,0.09901066621144612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,6144,0.09906399912304348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,6144,0.03871644536654154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,5120,0.08473866515689427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,65536,1.015933354695638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,5120,0.0337084432442983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,5120,0.08350577619340684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,4096,0.0734764469994439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,64,0.00902133352226681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,4096,0.06803555621041192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,3584,0.0618693364991082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,4096,0.06987733311123319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,3584,0.06018755833307902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,3584,0.06276977724499173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,10240,0.15946133931477866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,3072,0.05753066804673937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,3072,0.023520888553725347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,3072,0.05308444301287333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,3072,0.05568089087804159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,4096,0.028581334484948054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,2560,0.04580533173349169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,2560,0.02162844439347585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,2048,0.03547733359866672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,2560,0.048519111341900296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,7168,0.11133510536617702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,2048,0.03813777698410882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,2048,0.018519111805491976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,2048,0.04165244433614943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,2560,0.04642933275964525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,1536,0.0163102216190762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,1024,0.019693333241674636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,1536,0.031023999055226643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,1536,0.03447466757562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,768,0.016162667009565566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,1024,0.013565333353148567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,1024,0.02403733299838172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,1024,0.02784622377819485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,512,0.012817777693271637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,768,0.01984444426165687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,5120,0.08550844589869182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,768,0.024136000209384497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,256,0.012014221813943652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,512,0.017141333884663053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,512,0.019654222660594516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,8192,0.05052977800369263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,128,0.010435555544164447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,256,0.015206222732861837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,256,0.01585155559910668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,128,0.01107377807299296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,64,0.009694221946928237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,3584,0.0258542216486401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,32,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,1536,0.027131555808915034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,512,0.011649777491887411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,64,0.014483556151390076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,32,0.014683556225564746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,256,0.01074399964676963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,65536,0.7857679790920682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,128,0.014766222900814481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,768,0.012540444731712341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,65536,0.31469954384697807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,51200,0.6042622460259331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,16384,0.23203643163045248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,51200,0.25827378696865505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,16384,0.22114045090145537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,16384,0.20840266015794542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,12288,0.1537324455049303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,65536,0.8216844664679633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,51200,0.6640533341301812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,12288,0.06800711154937744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,10240,0.16304532686869302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,10240,0.13507021798027888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,51200,0.7287057770623101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,8192,0.11113954914940728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,10240,0.1353706651263767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,128,0.015531554818153381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,65536,0.898355589972602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,8192,0.042075554529825844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,7168,0.11087110969755386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,12288,0.15472888946533203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,7168,0.03826311230659485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,16384,0.08060089084837171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,6144,0.0992551114824083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,7168,0.09768088658650716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,10240,0.05285955468813578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,5120,0.08276800314585368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,6144,0.08388266960779826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,6144,0.08370400137371486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,12288,0.16245422098371717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,8192,0.10801955726411606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,5120,0.07139377461539374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,5120,0.029482665989134047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,5120,0.07280888822343615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,7168,0.09664711025026108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,4096,0.05870311127768623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,4096,0.06031733089023166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,6144,0.03409689002566867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,3584,0.05205600129233467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,3072,0.04961155520545112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,3584,0.05471555391947428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,4096,0.06382222308052911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,3072,0.04619733492533366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,3072,0.021130666136741638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,2560,0.041900446017583214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,3072,0.0486195551024543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,2560,0.019076445036464267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,2560,0.040074666341145836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,2048,0.033320890532599554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,2560,0.042528887589772545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,2048,0.01680088871055179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,2048,0.03358133302794562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,1536,0.026173333326975506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,2048,0.03659288750754462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,1536,0.014096889230940076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,1024,0.01753866672515869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,1536,0.030789332257376775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,1024,0.02124622298611535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,1024,0.024855999483002558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,768,0.014467555615637036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,1024,0.012145777543385824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,768,0.011416888899273343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,768,0.017314665847354464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,512,0.011333333121405708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,768,0.021540444758203294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,1536,0.027541332774692114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,256,0.008672888908121321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,512,0.0107351111041175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,512,0.01811022228664822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,256,0.01349866638580958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,256,0.010026666853162978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,128,0.007291555404663086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,256,0.015114666687117683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,3584,0.023423110445340473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,64,0.008639111287064022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,128,0.010114666488435533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,128,0.013110222087966071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,32,0.00887288898229599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,128,0.013529777526855469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,4096,0.02527911133236355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,64,0.01277955538696713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,32,0.013015111287434896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,512,0.015091554986106025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,65536,0.6635457674662272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,65536,0.24905334578620064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,51200,0.4658702214558919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,8192,0.10868889093399048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,65536,0.7471849123636881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,16384,0.16972266303168404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,51200,0.5861048698425293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,51200,0.19731911023457846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,3584,0.056326223744286426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,12288,0.13444356123606363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,16384,0.06530933247672187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,10240,0.11371021800571018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,12288,0.05144977900716993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,12288,0.13106400436825222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,12288,0.12812266084882948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,10240,0.043679111533694796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,8192,0.09182133277257283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,10240,0.10916977458530003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,8192,0.03556711143917508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,16384,0.17466755708058676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,8192,0.08912799755732219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,16384,0.1671919955147637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,65536,0.6826924218071831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,7168,0.03177511029773288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,7168,0.07962489128112793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,7168,0.07998755243089464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,10240,0.11104977793163723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,6144,0.02821777926550971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,6144,0.0696551137500339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,6144,0.07093066639370389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,8192,0.09007289012273152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,4096,0.0471955570909712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,5120,0.059083554479810924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,5120,0.06115644507937961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,4096,0.04884622163242764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,6144,0.07331911060545179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,3584,0.04155288802252875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,4096,0.05098844567934672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,3584,0.019753777318530612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,3584,0.0427706672085656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,3072,0.03735288977622986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,3584,0.04725955592261421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,7168,0.08610577715767755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,3072,0.01844888925552368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,2560,0.031786666976081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,3072,0.041598220666249595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,4096,0.021519111262427434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,2048,0.02606666584809621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,2560,0.03632888860172696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,5120,0.05844799677530924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,2048,0.028347555134031508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,2048,0.01480977733929952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,2048,0.03165422214402093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,5120,0.02510133385658264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,3072,0.03788266579310099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,1536,0.02351733379893833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,1536,0.013004444539546967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,1024,0.014937778313954672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,2560,0.01666844387849172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,1536,0.026649778087933857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,768,0.012395555774370829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,1024,0.011090666883521609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,1024,0.018159111340840656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,1024,0.02163022259871165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,512,0.009454222189055549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,768,0.010391111175219217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,768,0.015173332558737861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,768,0.018544889158672757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,256,0.007342221836249034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,512,0.012991999586423239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,51200,0.5191671053568522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,256,0.011799111134476133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,128,0.007335111498832703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,256,0.009133333133326637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,128,0.01144177797767851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,128,0.01311555587583118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,128,0.009203555683294931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,32,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,2560,0.03378666771782769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,64,0.01146666705608368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,256,0.01311377767059538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,32,0.011410666836632622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,1536,0.020263110597928364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,65536,0.5758488972981771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,512,0.009728000395827824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,65536,0.6695680088467069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,51200,0.45139646530151367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,512,0.015564444992277356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,16384,0.1484480036629571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,51200,0.18861867321862114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,65536,0.6668133205837674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,16384,0.060306668281555176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,12288,0.12499911255306667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,16384,0.16461866431766087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,51200,0.5059555371602377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,10240,0.10759199990166558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,12288,0.046999110115898975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,12288,0.12654844919840494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,51200,0.5052364667256674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,10240,0.10138399733437432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,8192,0.08869866530100505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,10240,0.10746133327484131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,16384,0.15663999981350368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,7168,0.07107555866241455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,8192,0.08193333281411065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,64,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,8192,0.08997511201434666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,12288,0.11834932698143853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,6144,0.06313066350089179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,7168,0.07216178046332465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,7168,0.07891466882493761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,6144,0.026732444763183594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,6144,0.07074133555094402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,5120,0.05382755398750305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,65536,0.23640267054239908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,5120,0.060658666822645396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,5120,0.023962666591008503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,7168,0.030192888445324365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,4096,0.045144889089796275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,4096,0.04961155520545112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,8192,0.03333688775698344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,4096,0.04676444331804911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,3584,0.019146665930747986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,3072,0.03573244478967454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,3584,0.038804445001814104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,3072,0.03558044301138984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,3584,0.039545777771208025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,3072,0.04110844598876105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,2560,0.029653333955340918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,2560,0.03149155444569058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,2560,0.035184890031814575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,2048,0.024703999360402424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,6144,0.06311644448174371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,2048,0.02608977754910787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,2048,0.030826667944590252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,1536,0.019941333267423842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,3584,0.0454711119333903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,1536,0.021793777743975323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,4096,0.02065333392884996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,1024,0.015305777390797934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,5120,0.05460355679194132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,1536,0.012536889149083031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,1536,0.02595022155178918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,1024,0.010753778119881948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,1024,0.016399110356966656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,1024,0.020855110552575853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,2048,0.014087110757827759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,768,0.014192000031471252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,768,0.017535110314687092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,512,0.012127999630239276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,2560,0.015966221690177917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,10240,0.04103999998834398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,256,0.007623111208279927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,512,0.009573332965373993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,768,0.012801777985360889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,256,0.01107377807299296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,256,0.008997333546479544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,128,0.006341333190600078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,256,0.012709333664841123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,768,0.010088000032636855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,64,0.007328888608349695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,128,0.009380444056457942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,128,0.010667555862002902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,128,0.012107555237081317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,64,0.01053066634469562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,512,0.01499644418557485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,65536,0.4525555504692926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,3072,0.017712000343534682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,512,0.009708444277445475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,65536,0.5606293148464626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,51200,0.4287048975626628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,32,0.010450666977299584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,51200,0.18010932869381377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,51200,0.4448560078938802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,16384,0.12515910466512045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,51200,0.39735110600789386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,16384,0.13650843832227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,16384,0.13122310903337267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,32,0.010399999717871347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,10240,0.09562577803929646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,12288,0.10288444492552017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,12288,0.1014488869243198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,65536,0.5093573464287652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,8192,0.07608266671498616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,10240,0.03769777880774604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,10240,0.08717777993943955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,16384,0.05520711011356778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,8192,0.02994489007525974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,8192,0.07158133056428698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,65536,0.22982933786180285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,12288,0.04053066505326165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,7168,0.0639537771542867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,6144,0.05775377485487196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,7168,0.0640995568699307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,10240,0.08756977981991237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,6144,0.025098666548728943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,5120,0.04721422327889336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,6144,0.05659021933873495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,12288,0.1095013353559706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,8192,0.072488890753852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,5120,0.021152888735135395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,5120,0.04883644315931532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,4096,0.0392213331328498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,4096,0.018722666634453666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,3584,0.03324800067477756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,4096,0.0413680010371738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,3584,0.017045333981513977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,6144,0.055735111236572266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,3072,0.030384000804689195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,3584,0.03762133253945245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,4096,0.037421332465277776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,3072,0.015602666470739575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,3072,0.03383377856678433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,7168,0.02658933401107788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,2560,0.02773244513405694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,2048,0.0209840006298489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,2560,0.014486221803559197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,3584,0.03526044554180569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,2048,0.023696889479955036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,2048,0.01274311128589842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,2048,0.026541334059503343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,5120,0.047500444783104785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,1536,0.01108266661564509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,1024,0.012140444583363004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,1536,0.021951110826598272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,2560,0.02990666694111294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,1024,0.014878221684032016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,1024,0.009722666607962714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,1024,0.017816889617178176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,1536,0.015047111445003085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,7168,0.05728533532884386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,512,0.009020444419648912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,768,0.012854222622182635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,768,0.00908088849650489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,1536,0.01963733302222358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,256,0.007332444190979004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,512,0.011129777464601727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,512,0.01313066648112403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,256,0.010072000324726105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,768,0.01035644445154402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,256,0.012038222617573209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,3072,0.0318524440129598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,768,0.01587466730011834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,128,0.012088889049159156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,128,0.008370666868156856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,512,0.00888888869020674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,32,0.006305777778228124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,64,0.009705777797434065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,32,0.009722666607962714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,65536,0.3914959960513645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,2560,0.026406221919589575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,65536,0.48538843790690106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,64,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,51200,0.37844355901082355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,65536,0.49994399812486434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,256,0.008368000388145447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,128,0.00666933341158761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,51200,0.37197422981262207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,16384,0.11248621675703262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,51200,0.14367822806040445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,12288,0.09209955400890774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,128,0.009689778089523315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,16384,0.11657510863410102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,51200,0.4105849001142714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,12288,0.03811111052831014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,12288,0.09040711323420207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,12288,0.10030577580134074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,10240,0.07545510927836101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,10240,0.03342311249838935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,10240,0.0860053300857544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,16384,0.04825777808825175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,16384,0.13163644737667507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,8192,0.027240888939963445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,7168,0.051739556921852954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,8192,0.07056266731686063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,7168,0.054936889145109385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,6144,0.04421955678198072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,7168,0.06456355253855388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,8192,0.06187911166085137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,6144,0.04799377918243408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,6144,0.022710222336981032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,6144,0.05593511131074694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,8192,0.061991108788384326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,4096,0.031242665317323472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,5120,0.049228443039788135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,7168,0.024850666522979736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,4096,0.03427022364404466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,4096,0.017650667164060805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,65536,0.1811119980282254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,3584,0.03074044320318434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,10240,0.07570311095979479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,3584,0.01635199950801002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,3584,0.037414222955703735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,3072,0.027637332677841187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,3584,0.027383110589451257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,2560,0.01998133295112186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,3072,0.01464000013139513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,5120,0.03913777735498216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,2560,0.01349244432316886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,5120,0.020284444093704224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,2560,0.0292595558696323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,2048,0.020790222618314955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,2048,0.025791111919615004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,4096,0.0415013333161672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,1536,0.013312888642152151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,1536,0.01716088917520311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,2560,0.024324445260895625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,1536,0.021380444367726643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,5120,0.04120799899101257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,3072,0.02334311107794444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,1024,0.009687110781669617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,1024,0.0131217779384719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,1024,0.017273777061038546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,3072,0.0336791111363305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,2048,0.012088889049159156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,512,0.00701333334048589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,768,0.011414222419261932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,768,0.00908799966176351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,768,0.015499555402331881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,256,0.006232888748248418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,512,0.00869511150651508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,512,0.010277333358923594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,2048,0.016543111867374845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,1536,0.010650667051474253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,128,0.005576000031497743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,256,0.008036444584528605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,256,0.009302222066455418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,64,0.0052675555149714155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,256,0.011436444189813403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,128,0.008199999729792276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,128,0.009025777379671732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,128,0.010815111299355825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,64,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,32,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,1024,0.009695111049546136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,512,0.012794666820102267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,768,0.00868711123863856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,65536,0.35880088806152344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,51200,0.31483999888102215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,51200,0.1129031048880683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,16384,0.09249955415725708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,51200,0.2830044428507487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,65536,0.13981066809760198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,16384,0.09800710943010117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,16384,0.038534220721986555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,16384,0.09563733471764459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,51200,0.29632266362508136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,32,0.005229333208666907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,10240,0.06024444765514798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,12288,0.07503022087944879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,65536,0.3266293207804362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,10240,0.0629991094271342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,10240,0.06443466742833455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,65536,0.3988328774770101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,12288,0.0703022215101454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,8192,0.05150755577617221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,7168,0.04148799843258328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,8192,0.021751999855041504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,7168,0.045921779341167875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,10240,0.02646844420168135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,7168,0.04856266578038534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,7168,0.020035554965337116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,6144,0.040172444449530706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,6144,0.04295555419392056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,5120,0.030561778280470107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,12288,0.029831111431121826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,12288,0.07432355483373006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,5120,0.016355555918481614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,4096,0.025799999634424847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,8192,0.04723644587728712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,5120,0.037680887513690524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,6144,0.038386666112475924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,3584,0.022870222727457683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,4096,0.014444443914625378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,4096,0.03230133321550157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,3584,0.02624711063173082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,3584,0.01329955541425281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,6144,0.018574222922325134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,3584,0.029313776228162978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,2560,0.01759466694460975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,3072,0.012087999946541257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,3072,0.023563555545277063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,5120,0.03471466567781236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,3072,0.02658577760060628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,2048,0.015021334091822306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,2560,0.011052444577217102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,2560,0.02091022166940901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,1536,0.012638222012254926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,2560,0.02367377777894338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,4096,0.029120888974931505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,2048,0.01773422294192844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,1024,0.009720000127951304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,1536,0.00903911143541336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,1536,0.014697778556081982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,1536,0.017436444759368896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,768,0.0081395556529363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,1024,0.011562666959232755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,1024,0.014496000276671516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,768,0.010266666611035665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,768,0.008052444292439355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,512,0.006823110911581252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,2048,0.010029333333174387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,2048,0.020435555113686454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,3072,0.020284444093704224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,512,0.011679111255539788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,8192,0.05309155583381653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,1024,0.008539555801285638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,128,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,256,0.008316444853941599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,256,0.010535111029942831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,768,0.013165333204799227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,64,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,128,0.010291555689440833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,32,0.004934222333961063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,512,0.009121778110663096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,256,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,64,0.008096888661384583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,32,0.008360000120268928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,256,0.007366221812036302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,65536,0.13345244195726183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,65536,0.37412712309095597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,51200,0.2780115604400635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,128,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,65536,0.3513760036892361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,51200,0.10734133587943183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,51200,0.28852711783515084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,51200,0.2757679886288113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,65536,0.2921431064605713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,12288,0.06608444452285767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,16384,0.09520710839165582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,12288,0.029440889755884807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,12288,0.07065955797831218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,128,0.008347555167145198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,10240,0.05682133303748237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,16384,0.08458222283257379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,8192,0.043040888177024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,10240,0.0603662199444241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,10240,0.06367999977535672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,16384,0.09250489208433364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,8192,0.049117333359188504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,8192,0.02152266601721446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,8192,0.052814222044414945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,7168,0.043712887499067515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,12288,0.07403288947211371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,512,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,6144,0.03875644339455499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,16384,0.03776622149679396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,6144,0.018236445056067575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,5120,0.030050665140151978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,5120,0.033687111404207014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,6144,0.03436977664629618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,5120,0.016193777322769165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,10240,0.026047999660174053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,4096,0.02836266822285122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,7168,0.03812444541189406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,4096,0.014283556077215405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,7168,0.019778667224778067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,4096,0.0314008891582489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,7168,0.04770044485727946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,4096,0.024467556013001338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,3072,0.019523556033770244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,3584,0.013096888860066732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,3584,0.0291608903143141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,3072,0.02286400066481696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,3072,0.01222222215599484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,2560,0.017343110508388944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,3072,0.026132444540659588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,2560,0.011185777684052786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,3584,0.02206666602028741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,5120,0.037759111987219915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,3584,0.02552266584502326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,2048,0.020152888364262052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,2048,0.009892444643709395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,1536,0.01276533305644989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,6144,0.04223733478122287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,1024,0.009672000176376766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,1536,0.009349333743254343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,1536,0.014304000470373364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,1536,0.016896888613700867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,768,0.00816177742348777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,1024,0.008369777765538957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,1024,0.014128888646761576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,2048,0.01700799994998508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,512,0.007244444555706448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,768,0.009927999642160203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,768,0.008176888856622908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,512,0.007638221813572778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,512,0.009045333498054082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,256,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,512,0.011316444310877057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,2560,0.02012622190846337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,2560,0.023227555884255305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,256,0.010748444332016839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,128,0.008001777860853408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,768,0.012875555290116204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,64,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,128,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,32,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,2048,0.015198222464985318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,1024,0.011173333558771344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,65536,0.2913057804107666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,65536,0.33042044109768337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,65536,0.13029600514305964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,64,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,32,0.007705777883529663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,51200,0.25757333967420787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,128,0.010372444159454769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,16384,0.07489244143168132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,51200,0.10213244623608059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,51200,0.27487466070387095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,16384,0.08271466361151801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,16384,0.035623110002941556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,128,0.004944889081848992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,16384,0.09451733032862346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,12288,0.028637333048714533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,10240,0.04178222351604038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,65536,0.3607235484653049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,10240,0.054142223464118115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,256,0.007336888876226213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,10240,0.025404444999165002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,10240,0.06336444616317749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,12288,0.06380266613430448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,8192,0.04457333352830675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,8192,0.02125777800877889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,8192,0.05243022243181864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,7168,0.04022666811943054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,256,0.008038221961922115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,6144,0.02719822194841173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,6144,0.03521600034501817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,7168,0.03536266750759549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,6144,0.01762488815519545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,5120,0.02568088968594869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,6144,0.04200889004601372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,51200,0.24019643995496961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,8192,0.0373324453830719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,5120,0.03064889046880934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,5120,0.03633244501219855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,12288,0.07365333371692233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,4096,0.025869333081775244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,3584,0.01902933253182305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,4096,0.013910222384664746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,3584,0.02308799988693661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,12288,0.05710488557815552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,3072,0.016576889488432143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,3584,0.028601778878106013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,3584,0.01293955577744378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,4096,0.020738666256268818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,2560,0.014851555228233337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,3072,0.011742221812407175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,3072,0.020936000678274367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,3072,0.025967111190160114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,2048,0.012798222402731577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,2560,0.01071733319097095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,2560,0.023179555932680767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,7168,0.047135111358430654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,7168,0.019706666469573975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,1536,0.009679999616410997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,2048,0.009763555394278632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,1536,0.013144888811641268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,1536,0.016866667403115165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,4096,0.03126666612095303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,1536,0.00908088849650489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,2048,0.0155404441886478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,1024,0.014102222190962898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,5120,0.015839111473825242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,1024,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,768,0.009343110852771336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,768,0.01309866706530253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,512,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,1024,0.010388444695207808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,2048,0.019688000281651814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,512,0.007681777907742395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,512,0.011347555451922946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,256,0.007679111427730984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,2560,0.017889777819315594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,1024,0.007640889121426477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,256,0.007339555356237624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,256,0.010391111175219217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,128,0.007349333001507654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,64,0.004264889078007804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,128,0.010079999764760336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,256,0.004975111120276981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,64,0.007357333269384172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,32,0.007335111498832703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,65536,0.166630228360494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,768,0.007977777885066139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,65536,0.12660888830820718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,512,0.008395555946562026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,65536,0.2870435449812147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,768,0.006772444479995304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,32,0.00424888895617591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,65536,0.3532177872127957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,16384,0.0607244438595242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,51200,0.10138755374484593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,51200,0.27572621239556205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,16384,0.07294399870766534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,12288,0.038895110289255776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,128,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,12288,0.05601244502597385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,10240,0.030787554052140977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,12288,0.07255288627412584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,51200,0.22230845027499727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,10240,0.04751911097102695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,8192,0.025694222913848028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,16384,0.09463200304243301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,8192,0.03939288854598999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,51200,0.19492888450622559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,12288,0.027906666199366253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,7168,0.02868355645073785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,8192,0.02053244411945343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,7168,0.018920888503392536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,7168,0.03562666641341315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,10240,0.024894222617149353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,6144,0.03147733211517334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,6144,0.04144622219933404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,6144,0.017157332764731515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,5120,0.01979466610484653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,5120,0.027278222971492346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,6144,0.023112000690566167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,4096,0.016728889611032274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,10240,0.0623555580774943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,5120,0.014927999840842353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,4096,0.01315199997689989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,4096,0.023065777288542852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,4096,0.030402666992611353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,8192,0.05211822191874186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,3584,0.020662221643659804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,3072,0.01349866638580958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,3584,0.027627555860413447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,5120,0.03598489032851325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,7168,0.04662400152948168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,2560,0.011956444217099084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,3072,0.011642666326628791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,3072,0.025286222497622173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,3584,0.014848000473446317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,2560,0.016232000456915963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,2560,0.010709332923094431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,2560,0.02200177808602651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,2048,0.009682666924264696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,2048,0.014038221703635322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,1536,0.00865955568022198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,2048,0.019273777802785236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,16384,0.03430577781465318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,1536,0.009030222064918941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,3072,0.018392889036072623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,1536,0.016537777251667447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,1024,0.008309333688682979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,768,0.005924444645643234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,1024,0.01384888920519087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,3584,0.012455110748608908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,768,0.0086666668454806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,512,0.005239999956554837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,768,0.01275644451379776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,128,0.004556444370084339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,1024,0.006607111129495833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,512,0.008013333711359236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,256,0.004565333326657613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,512,0.01108444482088089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,256,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,256,0.010002666877375709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,128,0.0042257776690853965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,768,0.008047111332416534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,1536,0.012100444071822695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,128,0.006955555743641323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,512,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,2048,0.010324444207880232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,64,0.0038773334688610504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,64,0.006928000185224745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,32,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,256,0.0073315559162033936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,65536,0.24345154232449004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,65536,0.08248444398244222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,65536,0.206840886010064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,1024,0.009655999640623728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,128,0.007334222396214803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,16384,0.05472266674041748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,51200,0.06706311305363973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,65536,0.16953778266906738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,16384,0.06243111027611626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,16384,0.023522666758961145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,16384,0.05904889106750488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,51200,0.15483910507626003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,12288,0.048341333866119385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,128,0.009742221898502773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,10240,0.03260533346070184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,12288,0.01982666717635261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,51200,0.18931999471452501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,8192,0.02530577778816223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,10240,0.04114400015936957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,10240,0.040694223509894475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,12288,0.03939022289382087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,8192,0.03459199931886461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,8192,0.03447644578086005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,7168,0.03133777777353922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,12288,0.046747555335362755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,6144,0.020043555233213637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,7168,0.013832000394662222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,10240,0.017941332525677152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,51200,0.16456711292266846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,5120,0.017525333497259352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,6144,0.012712000144852532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,6144,0.028535998529858057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,8192,0.014867555763986377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,4096,0.014740443891949125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,5120,0.023973332511054143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,5120,0.02416444487041897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,32,0.00424799985355801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,3584,0.013482666677898832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,7168,0.03124622172779507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,4096,0.020852444900406733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,3584,0.017779555585649278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,3072,0.01218488895230823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,3584,0.019395555059115093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,7168,0.02425066630045573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,5120,0.011756444142924415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,3072,0.016297777493794758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,2560,0.010784000158309937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,3072,0.017653332816229928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,4096,0.019842666056421068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,2560,0.00905600024594201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,2560,0.014462222655614218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,2560,0.016173332929611206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,3584,0.009776888622177972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,2048,0.008387555678685507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,1536,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,2048,0.014352000421947904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,1536,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,3072,0.009473777479595607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,1536,0.01275022245115704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,4096,0.010411555568377176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,2048,0.009304000271691216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,1024,0.00866844422287411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,1024,0.007436444362004598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,1024,0.011060444845093621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,2048,0.012501333322789935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,768,0.007102222078376346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,512,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,768,0.010277333358923594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,1536,0.010785777535703448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,512,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,512,0.009581333233250512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,256,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,256,0.00907733374171787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,256,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,128,0.003912000192536248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,6144,0.02757155564096239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,128,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,128,0.006305777778228124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,128,0.008795555267069075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,512,0.007329777710967594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,32,0.003560888684458203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,64,0.006464889066086874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,768,0.005592888842026393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,768,0.008013333711359236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,65536,0.20088177257113984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,64,0.003543111185232798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,1024,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,256,0.004132444245947732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,65536,0.0795653329955207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,65536,0.20048622290293375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,65536,0.1600231064690484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,51200,0.06361066632800631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,32,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,51200,0.1595591174231635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,12288,0.03021777669588725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,16384,0.023346667488416035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,16384,0.052503112289640635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,16384,0.056968887646993004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,12288,0.01940799918439653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,10240,0.023717333873112995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,12288,0.04688177837265862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,8192,0.02015644477473365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,10240,0.01780533293883006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,10240,0.03496888942188687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,16384,0.04232355621125963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,10240,0.03954666521814134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,7168,0.018724444839689467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,8192,0.014392000105645923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,51200,0.1598622269100613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,7168,0.013397333522637686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,7168,0.026628444592158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,7168,0.030312889152103003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,12288,0.04042133357789781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,6144,0.012441777520709567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,6144,0.02349511120054457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,51200,0.10910489161809285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,5120,0.011130666567219628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,4096,0.012372444073359171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,5120,0.023596444063716467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,8192,0.029558221499125164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,8192,0.03307377629809909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,3584,0.011156444748242697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,4096,0.017224889662530687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,4096,0.020244444410006206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,5120,0.014476444986131458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,3072,0.010040000081062317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,6144,0.02700533303949568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,5120,0.02015022271209293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,3072,0.014158222410413953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,2560,0.009015111459626092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,3072,0.01721244388156467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,4096,0.010110222631030613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,2560,0.012653333445390066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,2048,0.008085333638721043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,3584,0.015519110692871941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,3584,0.01847644481394026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,2048,0.011085333095656501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,2048,0.008370666868156856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,2048,0.014144889182514615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,3072,0.009354666703277165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,6144,0.01629333363638984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,1536,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,1536,0.012478222449620565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,2560,0.015448888142903646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,3584,0.009744889206356471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,1024,0.008012444608741337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,1024,0.007426666716734569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,1024,0.010739555789364709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,1536,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,512,0.004579555657174853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,768,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,768,0.010056888891590966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,1536,0.009672000176376766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,512,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,256,0.004233777936961916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,512,0.009655111365848118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,256,0.006614222294754452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,256,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,128,0.0038951109680864546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,256,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,128,0.006593777901596493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,128,0.00601244428091579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,128,0.00872444444232517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,32,0.003891555385457145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,1024,0.005863111052248213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,64,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,2560,0.00868622213602066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,65536,0.09119555685255264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,32,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,51200,0.08811644713083903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,65536,0.07811822493871053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,65536,0.15577867296006945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,65536,0.18978667259216309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,768,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,51200,0.12351911597781712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,51200,0.15023467275831434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,512,0.006753777878151999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,16384,0.02240266733699375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,16384,0.05405866768625048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,768,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,12288,0.033004442850748696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,10240,0.020873778396182593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,12288,0.04249155521392822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,51200,0.06251466936535306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,10240,0.028220444917678833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,12288,0.024136000209384497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,10240,0.01721955504682329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,8192,0.017595556047227647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,8192,0.023794665932655334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,12288,0.01862133377128177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,7168,0.01606666710641649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,8192,0.013809777796268463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,7168,0.02160888910293579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,16384,0.0417173339260949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,10240,0.037095109621683754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,7168,0.012788444757461548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,6144,0.018980445133315194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,8192,0.031034668286641438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,5120,0.012791111237472959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,6144,0.012081777883900536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,5120,0.016763556334707472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,4096,0.010825778047243753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,7168,0.028498666154013738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,6144,0.014342221948835584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,4096,0.010045333041085137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,4096,0.014150222142537435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,3584,0.009810667071077559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,4096,0.0191928893327713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,6144,0.025275554921891954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,3072,0.008755555583371056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,16384,0.033344000577926636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,3584,0.013108444710572561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,3584,0.017833777599864535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,2560,0.00776977754301495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,3072,0.00904888908068339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,3072,0.011785777906576792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,5120,0.011095110740926532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,3072,0.016192888220151264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,2048,0.0069422221018208405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,2560,0.010835555692513784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,2048,0.009711999860074785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,1536,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,64,0.0035457776652442086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,1536,0.00868711123863856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,1536,0.007675555845101674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,1024,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,2560,0.008845333423879411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,2048,0.008105778031879002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,1024,0.007163555257850223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,768,0.004646222210592694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,2048,0.013570666313171387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,1024,0.010594666832023196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,768,0.007019555403126611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,768,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,512,0.003919110943873723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,1536,0.012082666986518435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,512,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,512,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,512,0.009382222261693742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,256,0.005592888842026393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,256,0.008375999828179678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,256,0.006537777682145436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,5120,0.02202933364444309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,128,0.005618666609128316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,2560,0.014881778094503613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,256,0.0038942222793896994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,128,0.0064035554726918536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,32,0.003248888792263137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,3584,0.009464888936943477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,32,0.005620444400442972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,128,0.003589333345492681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,768,0.009712888962692684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,65536,0.15103644794887966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,65536,0.0776711106300354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,51200,0.06845955716239081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,128,0.00815644446346495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,64,0.005607999861240387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,51200,0.06213421954049004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,16384,0.02711822258101569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,65536,0.08943288856082493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,51200,0.14888266722361246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,12288,0.01958933307064904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,16384,0.022091555926534865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,16384,0.040878223048316106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,16384,0.053287999497519604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,10240,0.01719377769364251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,64,0.0032257777121331957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,65536,0.18838932779100206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,12288,0.04197955462667677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,8192,0.014723555909262763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,10240,0.016563556260532804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,10240,0.027830223242441814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,1024,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,51200,0.11980622344546848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,7168,0.013380444712109037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,8192,0.023330666952663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,8192,0.030385779009924993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,7168,0.0127688886390792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,6144,0.011930666863918304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,7168,0.027581334114074707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,12288,0.03234488765398661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,5120,0.010043555663691627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,6144,0.024567110670937434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,10240,0.03671111000908746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,5120,0.016369778248998854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,5120,0.011078221930397881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,7168,0.020928889513015747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,5120,0.02137600051032172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,12288,0.018124444617165458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,6144,0.0120000003112687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,4096,0.014179555906189812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,4096,0.010047111246320937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,4096,0.018613333503405254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,3584,0.009357333183288574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,3584,0.012822222378518848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,3072,0.0075866662793689305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,8192,0.01349688900841607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,3584,0.017250667015711468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,2560,0.00700711127784517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,3072,0.009015999734401703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,4096,0.00904266701804267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,3072,0.016230222251680162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,2048,0.006047111004590988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,6144,0.018558222386572096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,3584,0.008359111017651027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,2048,0.008101333346631791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,1536,0.005580444302823808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,2048,0.013175110850069257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,1536,0.008347555167145198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,1536,0.007640000018808577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,1536,0.011966221862369113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,2560,0.00870488915178511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,2560,0.014840889308187695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,1024,0.007342221836249034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,768,0.004242666479614046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,1024,0.010374222364690568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,2048,0.009702222214804756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,512,0.003931555483076307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,768,0.006984000404675801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,768,0.006648889018429651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,768,0.009655111365848118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,256,0.003536888708670934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,512,0.006657777975002925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,512,0.005998222364319696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,3072,0.011754666765530905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,512,0.00906577789121204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,256,0.006319111006127463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,2560,0.01070755554570092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,256,0.00870222184393141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,1024,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,64,0.0035546666218174826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,32,0.0032791110376516977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,128,0.008096000386608971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,65536,0.0588888857099745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,32,0.005629333357016246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,1024,0.004543999830881755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,65536,0.11286311679416233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,51200,0.046872887346479625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,256,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,65536,0.07772888739903767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,65536,0.18708444966210258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,16384,0.019274666905403137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,51200,0.06269333097669813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,51200,0.08862133158577813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,64,0.0053048887186580235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,16384,0.031026668018764917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,12288,0.01890222231547038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,16384,0.05275200141800774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,128,0.003573333223660787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,10240,0.016387555334303115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,12288,0.0414648883872562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,128,0.0063564446237352155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,10240,0.021270222134060327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,8192,0.01386044422785441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,10240,0.01645599967903561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,51200,0.14766755368974474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,16384,0.02115999990039402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,7168,0.01254666679435306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,8192,0.018016000588734944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,8192,0.030245333909988403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,12288,0.024752888414594863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,6144,0.010799110763602786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,7168,0.016536000702116225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,7168,0.027186666925748188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,128,0.00544177782204416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,5120,0.010050666828950247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,6144,0.014920888675583733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,6144,0.02409066590997908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,5120,0.01072622256146537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,8192,0.013487111363146039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,5120,0.021653332644038733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,12288,0.017940445078743827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,4096,0.011434666812419891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,3584,0.007882666256692674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,4096,0.018459555175569322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,6144,0.011824000212881299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,10240,0.03611911005443997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,3072,0.007069333559936947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,3584,0.017493334081437852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,5120,0.013358222113715278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,4096,0.008319999608728621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,2560,0.006607999818192587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,3072,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,3072,0.009707555174827576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,3072,0.01625066664483812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,7168,0.012757333616415659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,2048,0.005588444570700328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,2560,0.008617777791288164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,2560,0.014481777946154276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,3584,0.010738666686746808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,1536,0.004955555415815777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,2048,0.013412444127930535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,1536,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,1536,0.011787555283970304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,1024,0.004229333251714706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,4096,0.00979555563794242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,2560,0.00924622184700436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,1024,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,1024,0.0063164445261160536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,768,0.0042168887125121225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,1024,0.010168888502650792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,3584,0.009383999639087254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,512,0.0038897775941424896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,768,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,768,0.006004444426960415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,768,0.009670221971140968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,256,0.003531555748648114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,512,0.005619555711746216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,1536,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,512,0.009003555609120263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,128,0.003132444495956103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,256,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,256,0.008715555899673039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,128,0.0062977779242727495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,128,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,32,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,64,0.005208000126812193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,65536,0.03880533244874742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,512,0.006653333289755716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,256,0.0052800000541739995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,65536,0.1864480045106676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,65536,0.0773164431254069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,128,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,51200,0.08753599723180135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,51200,0.1475493378109402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,51200,0.061892443233066134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,65536,0.10888533459769355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,2048,0.00831733312871721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,16384,0.03010666701528761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,12288,0.012076444096035428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,16384,0.051968889103995435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,32,0.0052426668504873914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,12288,0.023743111226293776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,10240,0.012525333298577202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,12288,0.017652443713612027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,16384,0.014161777165200977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,10240,0.016401777664820354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,64,0.0028986665937635633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,8192,0.013167999684810638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,2048,0.008023999631404877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,16384,0.020977778567208186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,7168,0.012005333271291522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,8192,0.017733333839310538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,8192,0.02934755550490485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,12288,0.04020533296797011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,7168,0.01620533400111728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,7168,0.02681422233581543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,10240,0.020642666353119742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,10240,0.03500533435079787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,5120,0.009356444080670675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,6144,0.023762666516833838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,8192,0.01349688900841607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,5120,0.013055110971132914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,4096,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,5120,0.021021333005693223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,7168,0.012431999875439538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,6144,0.010710222025712332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,3584,0.007636444436179266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,4096,0.009705777797434065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,4096,0.01830044388771057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,51200,0.031015998787350122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,3072,0.006862222320503659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,3584,0.009389333426952362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,3584,0.01687733332316081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,3072,0.008901333643330468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,2560,0.006260444306664997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,3072,0.0158915552828047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,4096,0.011400889191362592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,2560,0.014494222071435718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,6144,0.014548444085650973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,2048,0.0052800000541739995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,6144,0.011728888584507836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,2048,0.007970666719807519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,1536,0.004888888862397936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,3072,0.009744000103738572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,2048,0.01311555587583118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,5120,0.01072533345884747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,1024,0.004217777815130022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,1536,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,1536,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,1536,0.011446221835083432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,1024,0.007072000039948358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,1024,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,2560,0.008419555922349295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,3584,0.01036977767944336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,768,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,2048,0.008356444537639618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,768,0.009694221946928237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,512,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,512,0.008671110702885522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,256,0.0031999999450312722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,1024,0.01037333326207267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,256,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,768,0.0069919998447100324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,128,0.003177777760558658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,256,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,128,0.004951111144489712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,128,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,128,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,64,0.004933333231343163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,32,0.0028684443483750024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,512,0.0033813334173626373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,65536,0.029751112063725788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,256,0.008039111064540016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,32,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,51200,0.02532355487346649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,65536,0.10805155833562215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,2560,0.009199110998047723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,64,0.0028515555378463534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,16384,0.010724444356229572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,51200,0.08715644147660996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,51200,0.14720977677239314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,16384,0.02983199887805515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,12288,0.012764444781674279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,16384,0.05168266759978401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,65536,0.07701422108544244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,12288,0.02357599967055851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,10240,0.011414222419261932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,12288,0.01774222155412038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,51200,0.061679999033610024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,10240,0.020632000433074105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,10240,0.016180444094869826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,10240,0.034683555364608765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,16384,0.02086488902568817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,8192,0.017439999514155917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,8192,0.02934044599533081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,768,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,7168,0.01626133256488376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,8192,0.010242666635248397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,7168,0.026457778281635706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,8192,0.013478221992651621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,6144,0.014542222023010254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,7168,0.009708444277445475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,5120,0.008347555167145198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,6144,0.011807999677128263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,6144,0.02370311154259576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,5120,0.013001778059535556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,5120,0.01074044406414032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,5120,0.020936000678274367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,6144,0.00907022257645925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,4096,0.009713778065310584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,4096,0.011411555939250521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,4096,0.017959111266665988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,512,0.006635555376609166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,3584,0.010383110907342698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,3072,0.006637333167923822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,3584,0.009352888498041365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,4096,0.007660444411966536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,3072,0.009420444567998251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,2560,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,3072,0.015495111544926962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,3584,0.007328000333574083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,2560,0.008947555389669206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,12288,0.040136890278922185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,3584,0.017126222451527912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,2048,0.008034666379292807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,2048,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,2048,0.013166222307417126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,2560,0.008364444805516137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,1536,0.007286222444640265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,65536,0.18653778235117593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,1024,0.003930666794379552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,1536,0.007678222325113084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,1536,0.011400000088744693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,7168,0.012503999802801343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,1024,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,1024,0.007135999699433644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,1536,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,768,0.005962666538026597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,768,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,768,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,2048,0.005421333428886201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,512,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,512,0.006577777779764599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,512,0.008667555948098501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,768,0.003887999802827835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,1024,0.010087110930018956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,128,0.002895999906791581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,256,0.006321777900060018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,256,0.00831200016869439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,512,0.0035573335157500375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,128,0.006304889089531369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,128,0.004979555391603046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,2560,0.014137778017255994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,32,0.002874666617976295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,256,0.0030986666679382324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,64,0.004966222163703707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,65536,0.030234667989942763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,51200,0.0244159996509552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,3072,0.008960000342792934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,16384,0.010039110978444418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,65536,0.10824800199932522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,12288,0.009030222064918941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,51200,0.08699822425842285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,64,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,16384,0.028982222080230713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,128,0.008056888977686564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,12288,0.023409777217441138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,7168,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,10240,0.020287111401557922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,6144,0.0073724447025193115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,8192,0.0174026671383116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,32,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,5120,0.007696888513035244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,7168,0.015969778100649517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,6144,0.01447111037042406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,5120,0.012786666552225748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,4096,0.011210666762457954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,3072,0.006305777778228124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,8192,0.008393777741326226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,2560,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,3072,0.00963644435008367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,2048,0.005078222188684675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,2560,0.008747555315494537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,256,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,1024,0.003899555653333664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,2048,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,768,0.003911111089918348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,1536,0.007146666447321574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,10240,0.00906133320596483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,1024,0.006303999986913469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,256,0.0028817777832349143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,768,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,128,0.0028693332440323303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,512,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,64,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,256,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,32,0.002569777684079276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,128,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,65536,0.02678044471475813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,64,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,51200,0.019194665882322524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,32,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,512,0.003548444559176763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,65536,0.10832533571455215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,3584,0.010399999717871347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,51200,0.08682222498787774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,10240,0.009359999663299983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,16384,0.02889333168665568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,8192,0.009023110899660323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,1536,0.004587555511130227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,4096,0.007293333609898885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,10240,0.020271110865804885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,6144,0.008104888929261101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,8192,0.01740977830357022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,3584,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,5120,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,4096,0.007531555162535773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,6144,0.014355555176734924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,5120,0.012803555362754397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,4096,0.011060444845093621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,3072,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,3584,0.010381333529949188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,2560,0.0057795556883017225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,3072,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,2048,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,2560,0.008798222574922774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,1536,0.004975111120276981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,2048,0.008048888709810045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,3584,0.007297777467303806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,1024,0.004229333251714706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,768,0.003907555507289039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,12288,0.02331377731429206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,7168,0.008756444685988957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,768,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,256,0.0032506665835777917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,7168,0.016163556112183463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,128,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,12288,0.010044444766309526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,64,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,128,0.004981333182917701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,1536,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,1024,0.006268444574541516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,64,0.004965333475006951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,32,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,16384,0.011751111182901593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,256,0.004976888911591636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,32,0.002576888849337896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,512,0.003371555565132035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,16384,1.0065778096516926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,512,0.005653333332803514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,16384,0.3982080088721381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,12288,0.6976497968037924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,16384,1.163816875881619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,12288,0.31059021419949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,12288,0.8192488882276746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,10240,0.5924711227416992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,10240,0.2633964485592312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,10240,0.6742106543646919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,8192,0.4651386472913954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,7168,0.4047351148393419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,8192,0.5399360126919216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,51200,3.805310991075304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,51200,1.3543368445502386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,6144,0.34462134043375653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,7168,0.18450755543178984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,51200,3.1779573228624134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,5120,0.3040124575297038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,6144,0.3959244357215033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,4096,0.24293600188361275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,8192,0.21072799629635283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,5120,0.1440782282087538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,3584,0.2653626600901286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,4096,0.11328800519307454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,3072,0.21938221984439424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,3584,0.23108887672424316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,2560,0.15273155106438532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,3072,0.09480177693896824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,5120,0.3354506757524278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,2560,0.07901244693332247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,2048,0.133715550104777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,2560,0.1695244444741143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,4096,0.2647537655300564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,1536,0.09570488664839004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,2048,0.0666746695836385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,3584,0.10462488730748494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,1024,0.07843110958735149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,1536,0.11009422275755142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,6144,0.1655786699718899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,3072,0.20022222730848524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,1024,0.082768890592787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,512,0.043453332450654775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,768,0.037744889656702675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,768,0.06735555330912273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,512,0.03529422150717841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,512,0.058655997117360435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,2048,0.1395137839847141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,256,0.03264710969395108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,128,0.026272889640596177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,768,0.057667553424835205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,128,0.03225955698225234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,64,0.022696000006463792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,128,0.04979822370741102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,64,0.04943733414014181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,256,0.03235110971662734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,7168,0.4638960096571181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,256,0.05373333228958977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,1024,0.0410755541589525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,32,0.023061333431137934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,32,0.04952088991800944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,1536,0.05217688944604662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,16384,0.8868328730265299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,65536,1.37390931447347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,51200,1.0489688449435763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,16384,0.8681484328375922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,51200,2.595532523261176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,16384,0.3237066798739963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,65536,3.369743982950846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,12288,0.24113154411315918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,12288,0.8470017645094129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,10240,0.2052675618065728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,51200,2.958715650770399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,10240,0.5167591306898329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,8192,0.4026159975263808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,7168,0.3782506783803304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,8192,0.4590017530653212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,7168,0.14903199672698975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,8192,0.16810754934946695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,6144,0.2849244541592068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,10240,0.46526045269436306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,5120,0.29813512166341144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,12288,0.7043422063191732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,4096,0.21161422464582655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,5120,0.11025333404541016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,4096,0.09154844284057617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,7168,0.3595297866397434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,4096,0.20870755778418648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,3072,0.17274578412373862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,6144,0.12780443827311197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,3584,0.18275645044114855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,2560,0.146053327454461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,3072,0.15745066271887884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,2560,0.06241866615083483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,3584,0.1696088843875461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,2048,0.12380445003509521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,3584,0.08279199732674493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,65536,3.684178670247396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,2048,0.11052622397740682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,3072,0.07055289215511747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,1536,0.04195377892918057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,1024,0.06553600231806438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,1536,0.08728266424602932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,2560,0.13389155599806044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,768,0.045997334851158984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,2048,0.05355022350947062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,1536,0.09652711285485162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,768,0.053949332899517484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,512,0.03924355573124356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,5120,0.26482311884562176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,256,0.02618666655487484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,512,0.029153777493370905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,512,0.04714488983154297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,128,0.02310133311483595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,1024,0.03353510962592231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,256,0.04304711023966471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,64,0.01960088809331258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,128,0.02666666607062022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,32,0.023175110419591267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,6144,0.31096090210808647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,64,0.039777778916888766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,768,0.03090755475891961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,32,0.03970400161213345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,256,0.02699822187423706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,1024,0.06612266434563531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,65536,0.44880978266398114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,128,0.04010488920741611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,51200,0.7366053263346354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,16384,0.285495122273763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,65536,1.1806631088256836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,51200,0.351872894499037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,16384,0.1091537806722853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,16384,0.2756551106770833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,10240,0.18660355938805473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,12288,0.21383821964263916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,8192,0.14024799399905735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,10240,0.0725839998986986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,10240,0.17197067207760283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,7168,0.12248710791269939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,8192,0.05770844221115112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,8192,0.13687733809153238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,6144,0.10245511266920303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,7168,0.12055022186703152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,12288,0.1856835550732083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,7168,0.05283200078540378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,5120,0.08399822314580281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,6144,0.10491733418570624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,4096,0.06751466459698148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,5120,0.040163556734720864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,12288,0.0830213361316257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,3584,0.06756177875730726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,4096,0.03303822212749057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,4096,0.07249333461125691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,65536,1.0770417319403756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,3584,0.06401689185036553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,2560,0.05485244592030843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,3072,0.05624710851245456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,6144,0.04513066675927904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,2560,0.023929778072569106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,2048,0.04450666573312548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,51200,0.9097937477959527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,1536,0.034322665797339544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,2048,0.02089333368672265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,2048,0.040355556541019015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,3072,0.05760888920889961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,1536,0.01771022213829888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,3584,0.029816001653671265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,1536,0.03289866778585646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,768,0.01943733294804891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,1024,0.015087111128701104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,512,0.01477688882086012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,1024,0.025391111771265667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,768,0.020990222692489624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,2560,0.04838755395677355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,512,0.017839999662505258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,5120,0.08859288692474365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,256,0.01186044431395001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,1024,0.023959999283154804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,256,0.015837333268589444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,64,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,128,0.011800000237094032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,32,0.009333333207501305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,768,0.01368622233470281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,64,0.015198222464985318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,32,0.015326221783955893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,256,0.01107911103301578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,128,0.010377777947319878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,65536,0.8031724294026693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,65536,0.3456097708808051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,51200,0.6577617857191298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,128,0.015488000379668342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,65536,0.9028746287027994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,51200,0.27497156461079914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,16384,0.08585955699284871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,12288,0.16663377814822727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,512,0.012583110895421771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,16384,0.22353243827819824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,10240,0.12409510877397324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,12288,0.06653155883153279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,8192,0.09862311018837823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,10240,0.058550225363837353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,12288,0.16564888424343532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,10240,0.1310746669769287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,8192,0.04782133301099142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,8192,0.10611022180981106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,3072,0.026650667190551758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,6144,0.08802488777372573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,16384,0.21704888343811035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,5120,0.07087910837597318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,6144,0.036814222733179726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,51200,0.6949973106384277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,4096,0.052348444859186806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,5120,0.06884177525838216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,7168,0.101419554816352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,4096,0.056608888838026256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,3584,0.05234666665395101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,3584,0.024521778027216595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,3072,0.04573511083920797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,6144,0.08101066615846422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,5120,0.032617777585983276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,3072,0.04381688767009311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,2560,0.020304888486862183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,2560,0.037905779149797224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,7168,0.04191644324196709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,2048,0.03194488750563727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,1536,0.024193776978386775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,3072,0.022673777408070032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,1536,0.025952888859642878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,1024,0.017681777477264404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,1536,0.01571999986966451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,2048,0.030997332599427965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,768,0.014871110518773397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,1024,0.013297778036859302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,1024,0.020587555236286588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,7168,0.09471378061506484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,512,0.012279999752839407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,768,0.01699911057949066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,256,0.00905777762333552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,512,0.014578666951921252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,2560,0.036526223023732506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,128,0.007337777978844113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,256,0.013097777962684631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,4096,0.026879999372694228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,64,0.007640889121426477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,128,0.012852444416946836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,768,0.012074666718641916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,2048,0.01827911039193471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,512,0.01110755569405026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,32,0.007689778175618913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,32,0.012784000072214337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,256,0.010513777534166971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,128,0.01070844464831882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,65536,0.6741591029696994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,51200,0.5093164443969727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,16384,0.18224533398946127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,51200,0.24069955613878039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,51200,0.5862017737494575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,64,0.0127688886390792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,12288,0.14677511321173772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,10240,0.12875466876559788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,12288,0.05805866585837471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,12288,0.13275467024909124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,8192,0.09522222148047553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,10240,0.11325244108835857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,3584,0.049750222100151904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,8192,0.040976001156700984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,8192,0.09093599849277073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,16384,0.08591022094090779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,7168,0.035840001371171736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,7168,0.08020799689822726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,6144,0.031812445984946355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,10240,0.05019911130269369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,5120,0.06535289022657606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,65536,0.7463528845045301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,5120,0.028513779242833454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,4096,0.050048000282711454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,7168,0.06877955463197497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,3584,0.03632355398601956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,4096,0.023987554841571387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,6144,0.07252889209323458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,3072,0.0384568903181288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,3584,0.0218551109234492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,3584,0.04327555497487386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,6144,0.07015022304322985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,3072,0.03815644317203098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,2560,0.01834488908449809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,5120,0.059663110309176974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,16384,0.19127378198835585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,4096,0.0488302210966746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,2048,0.028399109840393066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,65536,0.29144978523254395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,1536,0.023168888356950548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,1024,0.01440355512830946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,1536,0.014131555954615274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,2560,0.03348355491956075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,768,0.012162666353914471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,1024,0.01815822223822276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,768,0.011076444553004371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,512,0.010098666780524785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,3072,0.019987555013762582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,512,0.013129777378506131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,256,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,2048,0.025039111574490864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,1024,0.012138666378127204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,256,0.012058667010731168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,128,0.006600889066855113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,1536,0.020706666840447318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,768,0.015023110641373528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,64,0.005903999838564131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,128,0.01145066652033064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,2560,0.029098666376537744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,32,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,512,0.01037599974208408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,64,0.011439999772442712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,32,0.01142133358452055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,65536,0.5793537563747829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,256,0.00980799976322386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,65536,0.24096533987257215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,16384,0.14605244000752768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,65536,0.604489803314209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,51200,0.1863786644405789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,51200,0.4658800231085883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,16384,0.1445324420928955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,12288,0.10891022284825642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,2048,0.016559110747443307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,128,0.00998488896422916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,8192,0.07742044660780165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,10240,0.09229511022567749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,16384,0.06070666842990451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,7168,0.06615466541714139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,8192,0.03388177686267429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,8192,0.07491111093097262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,6144,0.05775644381841024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,7168,0.030511998467975195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,12288,0.04825333423084683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,7168,0.06748800145255195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,51200,0.37032355202568906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,6144,0.027271999253167048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,4096,0.03934311204486423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,5120,0.024544888072543677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,10240,0.041335999965667725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,3584,0.034544888469907976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,4096,0.020663999848895602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,4096,0.04161689016554091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,3072,0.030840890275107488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,3584,0.019095111224386428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,3584,0.0369137790468004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,2560,0.025703110628657874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,3072,0.017553778158293832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,3072,0.03308088911904229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,10240,0.08903377585940891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,2048,0.022664000590642292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,2560,0.016131555040677387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,6144,0.0583475563261244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,1536,0.01809955636660258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,2048,0.014599111345079211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,5120,0.04950755503442553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,2048,0.024490666058328416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,1536,0.01257777793539895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,12288,0.11128177907731797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,1024,0.010888000329335531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,1024,0.015494222442309061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,512,0.009316444396972656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,768,0.010217777556843227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,768,0.01331822243001726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,256,0.006672888994216919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,512,0.009518221848540837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,512,0.011404444773991903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,256,0.00903733323017756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,128,0.00665511108107037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,256,0.010429333481523724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,1024,0.014470222923490735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,128,0.009078222016493479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,768,0.011959999799728394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,64,0.010055111514197456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,32,0.010058666268984476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,5120,0.049550222025977246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,65536,0.47623109817504883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,2560,0.028763555818133887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,64,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,65536,0.22649065653483072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,51200,0.34206310907999676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,32,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,16384,0.12989777988857693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,12288,0.10153777731789483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,16384,0.05670577949947781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,16384,0.12689333491855198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,12288,0.044974221123589404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,10240,0.08297955327563815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,12288,0.09625777933332656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,8192,0.06784177488750882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,10240,0.08162844181060791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,1536,0.02011644509103563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,8192,0.03204177816708883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,8192,0.06646222538418241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,51200,0.1764213376575046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,51200,0.41873955726623535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,6144,0.04896266592873467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,7168,0.028986665937635634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,5120,0.04316888915167915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,6144,0.026016000244352553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,6144,0.05162577827771505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,4096,0.03384088807635837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,10240,0.04015288750330607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,5120,0.04374044471316867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,65536,0.532684432135688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,4096,0.019899555378490023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,3584,0.03059555424584283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,3584,0.01828177769978841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,3072,0.026152888933817547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,7168,0.05939733319812351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,3072,0.01664355562792884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,5120,0.023338667220539514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,2560,0.022602667411168415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,2560,0.015205333630243937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,128,0.010260444548394945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,2048,0.018792000081804063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,7168,0.05973066886266073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,1536,0.015040889382362366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,2048,0.013849777479966482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,2048,0.021877333521842957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,1024,0.011134222149848938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,1536,0.01219733307758967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,1536,0.018106665876176622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,1024,0.010290666586822933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,1024,0.013992889059914483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,512,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,768,0.009711110757456886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,768,0.012089778151777057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,256,0.006268444574541516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,512,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,4096,0.03654044535424974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,512,0.010635555618339114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,128,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,256,0.008715555899673039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,256,0.009700444837411245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,64,0.004967111266321606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,128,0.009015999734401703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,32,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,128,0.009703999592198266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,768,0.009176000124878353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,64,0.009425777528021071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,32,0.009706666900051964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,3072,0.029151999288135107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,65536,0.21386310789320204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,65536,0.4590889083014594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,16384,0.11031999852922227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,51200,0.35474665959676105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,16384,0.054339554574754506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,3584,0.03258400162061056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,16384,0.11388533645206028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,10240,0.07033066617117988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,12288,0.04095200035307143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,12288,0.08631733391020034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,65536,0.43639643987019855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,8192,0.05637600024541219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,51200,0.36364266607496476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,7168,0.0516115559471978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,51200,0.1492640044954088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,8192,0.0599822203318278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,6144,0.04359199934535556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,7168,0.025388444463411968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,7168,0.05287377701865303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,5120,0.03744622071584066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,6144,0.023168888356950548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,6144,0.046451555358039014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,4096,0.030271109607484605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,5120,0.020402666595247056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,5120,0.03935111231274075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,10240,0.03375377919938829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,3584,0.0270115551021364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,3584,0.01632355484697554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,12288,0.0837271081076728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,3584,0.029932445949978296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,3072,0.015014222926563688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,3072,0.026802667313151892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,10240,0.07281777593824598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,2560,0.023688889212078516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,2048,0.016164445214801364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,3072,0.02331999937693278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,2048,0.020211555891566806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,1536,0.013791110780504016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,4096,0.01773155563407474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,2560,0.013847110999955071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,1024,0.010659555594126383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,2048,0.013028444515334235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,1024,0.012783110969596438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,768,0.00904799997806549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,1024,0.009705777797434065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,4096,0.03310044606526693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,768,0.008995555341243744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,512,0.007181333170996771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,768,0.011422221859296164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,256,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,512,0.0086986662613021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,512,0.010034666293197209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,128,0.004906666775544484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,256,0.009120889008045197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,2560,0.019919999771647982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,128,0.008231110870838165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,64,0.004948444250557157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,32,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,8192,0.02972088919745551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,64,0.008707555631796518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,1536,0.01108977778090371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,65536,0.3963662253485785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,256,0.008347555167145198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,51200,0.29521332846747506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,128,0.008715555899673039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,2560,0.025807999902301367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,16384,0.08492889006932576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,32,0.008709333009190029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,51200,0.3148426744672987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,12288,0.0758133331934611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,16384,0.04639200038380093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,16384,0.09915733337402344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,10240,0.05619288815392388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,65536,0.4015928904215495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,1536,0.016367110941145156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,10240,0.06469688812891643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,8192,0.04771022333039177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,51200,0.13741778002844915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,7168,0.04517689016130236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,8192,0.026531555586391028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,8192,0.053146666950649686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,7168,0.02409244411521488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,7168,0.047188444270028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,12288,0.03663911090956794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,5120,0.032272001107533775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,6144,0.021939555803934734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,6144,0.04098666707674662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,4096,0.025964443882306416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,5120,0.019704888264338177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,10240,0.03247822324434916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,5120,0.035327110025617815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,4096,0.029715554581748113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,3072,0.020644444558355544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,3584,0.02698844505680932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,6144,0.03321866525544061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,3072,0.02419022222359975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,2560,0.016907556189431083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,65536,0.1800044510099623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,2048,0.01499022212293413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,2560,0.02123022245036231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,3584,0.02366222275627984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,4096,0.017204445269372728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,2048,0.012040889097584618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,2048,0.01789155602455139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,3584,0.015800888339678448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,1536,0.010597333312034607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,1536,0.014653333359294467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,768,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,3072,0.014297778407732645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,1024,0.011769777370823754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,512,0.006496888895829518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,768,0.008801778157552084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,768,0.01014666673209932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,256,0.0052800000541739995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,512,0.008365333080291748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,1536,0.013122667041089801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,128,0.004600888739029567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,256,0.008011555506123437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,256,0.008464888566070134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,64,0.004230222354332606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,128,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,128,0.008343111309740279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,1024,0.009415999882751042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,64,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,12288,0.07603200276692708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,32,0.008220444122950235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,2560,0.013232000172138214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,512,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,65536,0.324436452653673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,1024,0.00944622192117903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,16384,0.07566489113701715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,51200,0.10572266578674316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,32,0.004398222184843487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,12288,0.05989333656099108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,16384,0.08179822232988146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,12288,0.028584887584050495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,65536,0.259207116232978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,65536,0.1306177775065104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,10240,0.05290311243798998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,8192,0.040603554911083646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,51200,0.2523226737976074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,8192,0.021013332737816706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,7168,0.03326133224699233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,16384,0.03651288814014859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,6144,0.030239999294281006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,7168,0.02033244404527876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,10240,0.049921777513292104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,5120,0.0258577780591117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,12288,0.0624248915248447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,6144,0.01787644459141625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,6144,0.03450577788882785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,5120,0.016720000240537856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,5120,0.029663999875386555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,4096,0.013911111487282647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,4096,0.025068443682458665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,8192,0.04344888859324985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,3584,0.013111111190583972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,3584,0.023077333966890972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,7168,0.03907644417550828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,3072,0.012079999678664737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,3072,0.020502222908867728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,4096,0.021132444341977436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,51200,0.25232712427775067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,2048,0.012913777596420713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,2560,0.010899555351999072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,2560,0.01788444485929277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,1536,0.010703999963071613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,2048,0.010050666828950247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,2048,0.015341333217091031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,3072,0.016488000750541687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,1536,0.009480888644854227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,1536,0.01253600004646513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,768,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,1024,0.008646222452322641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,2560,0.014936889211336771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,768,0.008043555749787224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,10240,0.025497777594460383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,3584,0.019142222073343065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,512,0.008370666868156856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,256,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,1024,0.008648888932334052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,256,0.007694222033023834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,128,0.00461422238085005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,1024,0.010414222048388587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,512,0.006317333214812809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,768,0.009203555683294931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,32,0.004449777719047335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,512,0.007814221911960179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,65536,0.2597840097215441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,32,0.007674666742483775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,256,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,51200,0.21218755510118273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,65536,0.2902195453643799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,16384,0.05953866905636258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,51200,0.09910133149888779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,16384,0.03587822119394938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,16384,0.07293777995639376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,12288,0.05189244283570183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,64,0.007669332954618666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,10240,0.04257155458132426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,12288,0.05548355645603604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,65536,0.1273004478878445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,8192,0.03438399897681342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,10240,0.025390222668647766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,64,0.004401777767472797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,8192,0.021007110675175984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,7168,0.030660443835788306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,8192,0.03917510973082648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,6144,0.02527911133236355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,7168,0.01923199991385142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,7168,0.035079999102486506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,5120,0.020417778028382193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,6144,0.030849777989917334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,5120,0.016221332881185744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,10240,0.04698044392797682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,128,0.007686222592989604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,4096,0.013678222066826291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,3584,0.016751999656359356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,6144,0.017495110630989075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,3584,0.020512888828913372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,3072,0.014837332897716098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,5120,0.026583110292752583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,51200,0.22473510106404623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,2560,0.013076444466908773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,4096,0.022709333234363135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,2560,0.015967110792795818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,12288,0.028155555327733357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,3584,0.012801777985360889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,2048,0.013748444616794586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,1536,0.009348444640636444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,128,0.007476444045702617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,3072,0.01827822294500139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,1024,0.007295110987292395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,1536,0.00904266701804267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,1536,0.011409777734014722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,768,0.006275555739800136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,2560,0.011074666347768573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,1024,0.009697777529557547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,512,0.0052435555391841466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,768,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,768,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,256,0.004606222112973531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,512,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,512,0.007914666500356462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,128,0.004231111043029361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,256,0.007293333609898885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,3072,0.011763555308183035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,256,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,128,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,32,0.004046222401989831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,128,0.007364444434642792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,2048,0.011386666860845355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,64,0.007038222418891058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,65536,0.23352710405985513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,32,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,51200,0.18491200606028238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,2048,0.009759999811649323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,65536,0.27111999193827313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,16384,0.0665644473499722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,51200,0.21075467268625894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,64,0.0038924444880750445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,12288,0.051035556528303355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,16384,0.03513155712021722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,12288,0.027806222438812256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,12288,0.053103112512164645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,65536,0.12177511056264241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,10240,0.04520977867974175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,8192,0.03435111045837402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,16384,0.06931822167502509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,8192,0.03738222188419766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,8192,0.020527111159430612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,10240,0.0414737794134352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,4096,0.018409777018758986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,7168,0.018696000178654987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,6144,0.025775111383861963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,10240,0.02481599979930454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,5120,0.02168088820245531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,6144,0.029899554120169744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,1024,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,4096,0.018255111243989732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,5120,0.025820444027582806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,7168,0.02968888812594944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,3584,0.016538666354285348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,4096,0.02215822206603156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,3072,0.0148026661740409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,7168,0.033942222595214844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,3584,0.020049777295854356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,2560,0.01309511148267322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,3072,0.017666666044129264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,2560,0.01073333372672399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,2048,0.011397333608733283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,4096,0.013916444447305469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,2048,0.013447110851605734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,3584,0.012479110724396177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,1536,0.009169778062237633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,51200,0.09840977854198879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,1024,0.006991110742092133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,1536,0.011332444846630096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,5120,0.01574488812022739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,768,0.0063991112013657885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,2048,0.009874666730562845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,768,0.008283555507659912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,3072,0.011624000138706632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,512,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,512,0.007621333003044128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,2560,0.01551555593808492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,512,0.007632888853549957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,1024,0.009366222553782994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,128,0.004269333349333869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,256,0.007027555671003129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,768,0.008340444829728868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,64,0.004202666795916027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,128,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,32,0.003936000168323517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,64,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,1024,0.008356444537639618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,256,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,65536,0.16943466663360596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,256,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,51200,0.1320657730102539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,65536,0.12024711238013373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,16384,0.050492445627848305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,65536,0.22738578584459093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,128,0.007324444750944774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,12288,0.03844889005025228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,6144,0.017104888955752056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,16384,0.05967555443445841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,12288,0.02678311036692725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,12288,0.04570666617817349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,10240,0.02402488887310028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,10240,0.03945422172546387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,51200,0.09533600012461345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,8192,0.032949331733915545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,1536,0.008995555341243744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,8192,0.019679110911157396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,10240,0.031818668047587075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,7168,0.01791288952032725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,6144,0.02015822297996945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,7168,0.029412445094850328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,5120,0.01739022301303016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,6144,0.016211556063758004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,8192,0.026102221674389307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,6144,0.02632355524433984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,4096,0.015421332584487068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,5120,0.015180443723996481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,5120,0.022652443912294176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,4096,0.01276711126168569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,51200,0.179312891430325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,4096,0.018969777557584975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,7168,0.02253066665596432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,3584,0.012038222617573209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,3584,0.017640888690948486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,2560,0.011391999820868174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,32,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,3072,0.015640000502268474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,2048,0.009363555245929295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,2560,0.013757333159446716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,1536,0.007648888561460707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,2048,0.012075555821259817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,3584,0.013596444494194455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,1536,0.010151111417346531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,1024,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,3072,0.012068444656001197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,3072,0.01127644462717904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,768,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,1024,0.008642666869693333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,2560,0.01035644445154402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,512,0.00489066665371259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,768,0.007991111112965478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,768,0.007702222300900354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,256,0.004249777644872665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,512,0.007297777467303806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,16384,0.03300177719857957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,128,0.003583999971548716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,1024,0.008185777399275038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,128,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,64,0.003536000019974179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,2048,0.009597332941161262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,32,0.0035333335399627686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,512,0.007635555333561367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,64,0.006640000061856375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,65536,0.1977848874198066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,256,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,51200,0.1409955554538303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,128,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,51200,0.1496702167722914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,16384,0.04881333311398824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,16384,0.023424888650576275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,12288,0.03677599959903293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,1536,0.009010666774378883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,12288,0.019107555349667866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,12288,0.03840533230039809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,10240,0.02975377771589491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,51200,0.061347557438744434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,10240,0.017500445246696472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,8192,0.025091555383470323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,8192,0.014144000079896716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,16384,0.049584001302719116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,8192,0.027747554911507502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,256,0.007230222225189209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,6144,0.01941244469748603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,7168,0.01311466677321328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,7168,0.02516533268822564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,5120,0.01680177781316969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,6144,0.012030222349696688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,6144,0.022650667362742957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,4096,0.0144533332851198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,5120,0.011420444481902652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,10240,0.0329022208849589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,5120,0.019277332557572257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,4096,0.01036977767944336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,4096,0.01611911091539595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,7168,0.021976888179779053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,3584,0.009713778065310584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,3072,0.009374221993817223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,65536,0.07758133278952704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,2560,0.01070844464831882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,2560,0.008857777549160851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,2048,0.009006222089131674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,3584,0.013148444394270578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,2048,0.008384889198674096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,3072,0.011889778077602386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,65536,0.19030400117238364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,3072,0.013540444274743399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,1536,0.009015111459626092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,1536,0.008011555506123437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,2560,0.012111999922328524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,1024,0.007292444507280986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,1024,0.007683555285135905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,768,0.00723555518521203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,512,0.004573333180612988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,32,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,512,0.006731555693679386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,512,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,256,0.003927111211750242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,256,0.006671111202902264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,768,0.005565333283609814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,2048,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,128,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,1536,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,64,0.003540444291300244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,128,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,32,0.0035235554807715942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,65536,0.14569955401950413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,1024,0.006375999914275275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,32,0.006015111174848344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,51200,0.1104746659596761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,128,0.003964444415436851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,65536,0.18215999338361952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,16384,0.037484443849987455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,51200,0.05928444200091892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,768,0.007087111473083496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,3584,0.01482399966981676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,12288,0.02635555631584591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,16384,0.022280888424979314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,16384,0.04736622174580892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,10240,0.022475555539131165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,12288,0.018541332748201158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,64,0.005980444451173146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,10240,0.01684533390733931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,10240,0.03199377655982971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,65536,0.07300533188713922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,256,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,7168,0.01889600025282966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,8192,0.013555555707878537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,8192,0.02700088918209076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,7168,0.012523555921183692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,51200,0.1423164473639594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,5120,0.014581332604090372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,6144,0.021339555581410725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,5120,0.010763555765151978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,12288,0.037236445479922824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,4096,0.012588444683286877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,4096,0.009935110807418823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,3584,0.011438222395049201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,6144,0.016774222254753113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,7168,0.023965333898862202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,3584,0.01424888935354021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,6144,0.011583999627166323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,3072,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,2560,0.008951110972298516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,8192,0.021164443757798936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,5120,0.018242667118708294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,2560,0.008718222379684448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,4096,0.015467555986510383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,2048,0.010408889088365767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,2048,0.008403555386596257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,3584,0.00960000024901496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,1536,0.00793333351612091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,3072,0.010091555615266165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,1024,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,3072,0.01312711089849472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,2048,0.007689778175618913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,768,0.006949333681000604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,768,0.007015110717879401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,1536,0.006700444552633498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,512,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,1024,0.005842666659090254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,768,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,256,0.006162666612201267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,128,0.003544888976547453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,256,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,2560,0.011762667033407422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,64,0.0032248888164758682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,128,0.006463999963468975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,128,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,32,0.0032177778581778207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,65536,0.09592444366878933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,512,0.00654666663871871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,32,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,65536,0.07254133621851604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,1024,0.007678222325113084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,65536,0.13960088623894587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,16384,0.028916445043351915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,51200,0.05800977680418226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,512,0.004560889055331548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,12288,0.026854222019513447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,16384,0.038197331958346896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,64,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,1536,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,10240,0.023054222265879314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,12288,0.029976887835396662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,8192,0.01941155559486813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,51200,0.08057689004474215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,10240,0.026029333472251892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,7168,0.01759199963675605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,8192,0.021374222305085924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,51200,0.10941333240932888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,7168,0.01238844460911221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,7168,0.019446222318543326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,256,0.003913777983850903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,12288,0.01756088932355245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,6144,0.011352000137170156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,6144,0.01752088963985443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,4096,0.010663111176755695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,5120,0.010415111151006486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,5120,0.015207111835479736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,3584,0.010422222316265106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,4096,0.01312622262371911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,3584,0.009339555270142024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,6144,0.014020444618331062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,3072,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,3072,0.009176000124878353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,5120,0.012089778151777057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,3072,0.011169777976142036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,10240,0.015970667203267414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,2560,0.008599999878141616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,2048,0.007303111255168915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,2560,0.010103111465771994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,8192,0.013455111119482251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,1536,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,2048,0.008237332933478886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,1024,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,1536,0.007670222057236566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,16384,0.02163377735349867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,3584,0.012236444486512078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,1024,0.007090667055712805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,1024,0.006989333364698622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,512,0.004233777936961916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,768,0.006993778049945831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,768,0.006447111152940326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,512,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,512,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,2048,0.009184000392754873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,1536,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,256,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,128,0.0035342222286595237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,128,0.006322666588756774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,64,0.0032151111712058387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,128,0.005625777774386936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,32,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,4096,0.009677333136399588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,768,65536,0.07116978036032783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,256,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,65536,0.07255466779073079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,65536,0.13560977247026232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,51200,0.05756711297565036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,51200,0.10500978098975287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,256,0.003584889074166616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,32,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,16384,0.020804444948832195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,16384,0.03648799988958571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,768,0.0047173334492577445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,12288,0.017122666041056316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,768,51200,0.05202755663130018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,12288,0.029091556866963703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,768,8192,0.015224888920783997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,10240,0.015903110305468243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,10240,0.024688000480333965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,768,16384,0.023402666052182514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,64,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,7168,0.012303999728626676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,7168,0.018594667315483093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,768,6144,0.012171555724408893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,768,10240,0.01777866648303138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,6144,0.011406222151385413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,768,5120,0.010976889067225985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,6144,0.01683111157682207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,768,4096,0.009338666995366415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,5120,0.010416000253624385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,5120,0.014856000741322836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,8192,0.013186666700575085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,768,3584,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,4096,0.012848888834317526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,768,3072,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,3584,0.009356444080670675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,3584,0.01180355581972334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,3072,0.009000000026490953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,3072,0.010955555571450127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,2560,0.008004444340864817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,768,2048,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,2560,0.009992000129487779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,768,7168,0.013614222407341003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,768,1536,0.0052826665341854095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,768,12288,0.019719999697473314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,1536,0.007666666474607255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,768,2560,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,1536,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,1024,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,2560,0.008552889029184977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,768,768,0.004249777644872665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,1024,0.0068408888247278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,2048,0.008082666330867344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,768,0.006835555450783835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,512,0.006670222100284364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,512,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,768,256,0.003549333247873518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,768,1024,0.004593777573770947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,768,128,0.0032426667296224167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,256,0.005591111050711737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,128,0.0063271112740039825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,768,64,0.002862222285734283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,768,512,0.003726222034957674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,2048,0.008974221845467886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,64,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,512,65536,0.049978666835361056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,32,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,256,0.006331555545330048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,512,51200,0.03924444317817688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,65536,0.09603110949198405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,128,0.005305777821275923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,51200,0.057368887795342334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,51200,0.07458755705091688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,4096,0.009524444739023844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,512,12288,0.017246221502621967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,16384,0.020852444900406733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,16384,0.026875555515289307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,512,10240,0.01588444411754608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,12288,0.017237333787812125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,12288,0.021672000487645466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,65536,0.07244266404045953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,10240,0.01588000026014116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,10240,0.01861422260602315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,512,7168,0.012811555630630918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,512,16384,0.017712000343534682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,8192,0.020637333393096924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,7168,0.01483022173245748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,6144,0.013454222016864352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,512,8192,0.016145777371194627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,6144,0.011234666738245221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,512,5120,0.009680888719028896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,8192,0.015954666667514376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,768,32,0.0029146667155954572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,7168,0.012079111403889127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,5120,0.012063999970753988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,4096,0.009384888741705153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,4096,0.010415111151006486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,3584,0.010048888623714447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,3584,0.00904266701804267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,512,3072,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,512,6144,0.010814222196737925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,3072,0.008725333544943068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,512,3584,0.008094222181373173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,2560,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,512,2048,0.005592888842026393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,768,0.006319111006127463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,512,4096,0.008599999878141616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,512,1536,0.005241777747869492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,2048,0.007713778151406183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,1536,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,1536,0.006690666493442323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,2560,0.008419555922349295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,1024,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,512,768,0.003929777691761653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,1024,0.007154666715198093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,5120,0.01037333326207267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,512,512,0.003551111039188173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,2048,0.008004444340864817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,768,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,512,256,0.0032755554550223877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,512,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,512,1024,0.004219555606444676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,512,128,0.0028728888266616394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,256,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,256,0.005315555466545953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,8192,0.013136000268989138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,128,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,512,32,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,768,0.0070426662762959795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,256,65536,0.032015111711290144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,512,2560,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,512,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,256,51200,0.02886755598915948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,65536,0.07129244671927558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,65536,0.09344089031219482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,256,16384,0.01275644451379776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,512,64,0.0028826666788922418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,51200,0.05606844690110949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,51200,0.07344266441133288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,256,12288,0.011976000335481433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,16384,0.0203155560625924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,16384,0.025976889663272437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,12288,0.017158221867349412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,32,0.005234666582610872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,256,8192,0.010407110883129967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,10240,0.018391110830836825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,8192,0.015672889020707872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,256,7168,0.00999022192425198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,256,10240,0.011101333631409539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,12288,0.020971554848882887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,10240,0.015803555647532146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,7168,0.012091555529170565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,6144,0.011334222224023608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,256,5120,0.009332444104883406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,128,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,5120,0.010144888526863521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,5120,0.011800888511869641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,64,0.004931555440028508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,7168,0.014477334088749357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,256,3584,0.007280888656775157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,4096,0.009392889009581672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,3584,0.009006222089131674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,256,3072,0.006611555400821898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,3072,0.009018667042255402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,256,4096,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,256,2560,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,3072,0.00870400004916721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,256,6144,0.010040889183680216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,4096,0.010436444646782346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,256,2048,0.00554666668176651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,2560,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,256,1536,0.004559999952713649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,2048,0.008007110820876228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,3584,0.009744000103738572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,256,1024,0.004210666649871402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,1536,0.007431999676757389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,1536,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,256,768,0.0035546666218174826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,1024,0.006170666466156642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,768,0.006950221955776215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,768,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,2560,0.008332444561852349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,512,0.006640000061856375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,8192,0.012930666406949362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,2048,0.007671111159854465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,256,0.0063235556913746735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,1024,0.007017778025733099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,256,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,128,0.006342222293217977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,128,0.004908444566859139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,256,64,0.002862222285734283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,6144,0.01311377767059538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,64,0.004912000149488449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,256,128,0.0029333333174387612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,32,0.004958222309748332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,128,65536,0.02488800055450863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,256,512,0.0032151111712058387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,128,51200,0.0185208883550432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,65536,0.09313333034515381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,128,16384,0.01036444471942054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,51200,0.05493422349294027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,256,256,0.0032097777972618737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,256,32,0.0030417777597904205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,16384,0.019951999187469482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,16384,0.025851555996470984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,128,10240,0.011629333098729452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,12288,0.016875555117925007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,12288,0.02073777715365092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,128,8192,0.01037688884470198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,10240,0.015651555524932016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,128,7168,0.009738666315873465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,8192,0.012818666795889536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,8192,0.01551644504070282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,128,6144,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,512,0.005624889085690181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,51200,0.07447377840677898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,128,5120,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,6144,0.011072888970375061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,5120,0.010075555907355415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,128,4096,0.007698666718271043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,10240,0.01822933389080895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,4096,0.010113777385817634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,128,3584,0.006959999601046245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,7168,0.01441422270403968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,128,12288,0.009034666750166152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,128,3072,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,6144,0.01309866706530253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,3584,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,128,2560,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,3072,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,5120,0.011723555624485016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,128,2048,0.005249777601824866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,2560,0.008347555167145198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,65536,0.07101333141326904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,2560,0.008314666648705801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,2048,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,4096,0.009369778136412302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,2048,0.007675555845101674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,1536,0.0074106670088238185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,1536,0.00666933341158761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,128,768,0.0036106668412685394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,1024,0.0069822221994400024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,3584,0.009053332938088311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,128,512,0.003277333246337043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,768,0.006725333217117522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,768,0.0057706667317284485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,3072,0.00903911143541336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,128,256,0.0029724445194005966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,512,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,128,128,0.0028862222615215513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,256,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,7168,0.011770666473441653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,128,1024,0.003900444342030419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,128,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,128,32,0.0028657778683635923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,128,0.004972444640265571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,64,65536,0.02405599918630388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,1024,0.005976888868543837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,64,51200,0.015999999311235216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,512,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,64,16384,0.011448889142937131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,51200,0.07413955529530843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,256,0.0052328887912962176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,64,12288,0.008722667064931657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,64,10240,0.00831822223133511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,16384,0.0255724440018336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,64,8192,0.00832444429397583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,12288,0.020640000700950623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,64,7168,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,10240,0.017986666825082567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,64,6144,0.007354666789372762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,65536,0.09276266892751057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,7168,0.014156444205178155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,128,1536,0.004679999831649992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,64,4096,0.007671111159854465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,6144,0.012812444733248817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,128,64,0.002862222285734283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,5120,0.011643555429246692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,64,3072,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,4096,0.010094222095277574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,64,0.005166222237878376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,32,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,64,2048,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,8192,0.015493333339691162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,64,1536,0.004550222307443619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,64,5120,0.00868444475862715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,64,1024,0.004003555410438114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,2048,0.007335111498832703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,1536,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,64,768,0.00370399985048506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,1024,0.006095110956165526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,64,512,0.003552888830502828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,64,2560,0.005631111148330901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,64,256,0.0028880000528362063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,3072,0.00903733323017756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,64,128,0.0028764444092909494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,512,0.0052826665341854095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,2560,0.008210666477680206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,256,0.005299555758635203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,64,32,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,128,0.004947555561860402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,32,65536,0.018995554910765756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,32,0.004973333328962326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,32,51200,0.015133332875039844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,3584,0.009640889035330879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,768,0.005762666463851929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,65536,0.09240800142288208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,32,12288,0.014099554883109199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,51200,0.07289599710040622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,64,64,0.0029288888391521242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,16384,0.02551377813021342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,32,8192,0.009342222577995723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,12288,0.020639111598332722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,64,3584,0.0069573331210348345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,10240,0.018010665973027546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,32,6144,0.008362666600280339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,8192,0.015503111812803479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,32,5120,0.012401777837011548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,32,4096,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,32,16384,0.01719199948840671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,32,3584,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,5120,0.011567999919255575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,32,10240,0.009394666386975182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,4096,0.010034666293197209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,32,2560,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,3584,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,32,2048,0.005314666777849197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,64,0.004944889081848992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,3072,0.008999110923873054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,32,1536,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,32,1024,0.0038986665507157645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,2048,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,32,768,0.0035377778112888336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,7168,0.014274666706720987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,32,512,0.0031795555518733132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,1024,0.006250666661394968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,32,256,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,768,0.005607999861240387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,32,3072,0.006953777538405524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,512,0.005478222337034013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,32,64,0.00254844439526399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,256,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,32,32,0.0028764444092909494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,32,7168,0.019271110494931538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,128,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,2560,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,64,0.004951111144489712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,32,0.004964444372389051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,1536,0.006664888726340399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,6144,0.013092444174819522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,769,32,128,0.0028213332924577924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,16384,0.9978497823079427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,16384,1.0141635470920138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,12288,0.6842248704698352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,51200,1.0903324551052518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,12288,0.7336186832851834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,51200,2.9724604288736978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,16384,0.3340880076090495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,12288,0.29140178362528485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,10240,0.5582462416754829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,51200,3.5706570943196616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,10240,0.21932444307539198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,8192,0.45905065536499023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,10240,0.6130551232231988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,12288,0.65756713019477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,8192,0.4905502001444499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,16384,0.9053804609510633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,7168,0.3864266607496474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,10240,0.5549271371629503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,8192,0.17243999905056426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,7168,0.15723289383782282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,8192,0.45227734247843426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,6144,0.13844088713328043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,7168,0.3840924368964301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,5120,0.31151821878221303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,6144,0.332514656914605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,4096,0.22088532977634004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,5120,0.11845866839090984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,7168,0.4202711052364773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,6144,0.37035467889573837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,5120,0.28009867668151855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,4096,0.24348444408840605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,3584,0.2093555529912313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,4096,0.23652089966668022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,3584,0.0869706670443217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,5120,0.3033217853969998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,6144,0.331371545791626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,3584,0.2139528857337104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,4096,0.09707644250657822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,3072,0.0746871100531684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,3072,0.18008089065551758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,2560,0.18181867069668242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,51200,3.0491227043999563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,2560,0.15865777598487005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,2560,0.06698311037487455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,3072,0.21173600355784097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,2560,0.15281777911716038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,3584,0.20192710558573404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,1536,0.10625688897238837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,2048,0.13094311290317112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,2048,0.12433066633012559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,1024,0.0731582244237264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,1536,0.10309333271450466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,1536,0.09938666555616592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,1024,0.07832711272769503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,1024,0.03647022114859687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,2048,0.1374666690826416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,768,0.05338311195373535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,2048,0.05801066425111559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,768,0.06401333543989393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,3072,0.18583288457658556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,1536,0.04481866624620226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,256,0.02942311101489597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,512,0.05541511045561897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,512,0.030213332838482324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,512,0.05202044381035698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,128,0.02478666603565216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,256,0.02787999974356757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,1024,0.07694222529729207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,256,0.04192177785767449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,128,0.02772177755832672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,128,0.046909332275390625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,512,0.04098133246103922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,32,0.02156977852185567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,64,0.04650488826963636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,32,0.04630400074852837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,256,0.05067466696103414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,64,0.021446221404605444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,128,0.03427822060055203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,768,0.032941334777408175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,768,0.06274222003089057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,65536,1.0346097946166992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,65536,3.212429258558485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,51200,2.534133275349935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,65536,3.4010630713568797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,16384,0.9522755940755209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,51200,2.7677555084228516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,51200,0.8235528733995227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,16384,0.26127555635240346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,16384,0.804732428656684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,12288,0.730438232421875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,12288,0.5757617950439453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,10240,0.48340712653266055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,16384,0.6996791097852918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,12288,0.5134542253282335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,12288,0.19758844375610352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,10240,0.4300373395284017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,8192,0.3876053227318658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,8192,0.1378799941804674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,7168,0.31513688299391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,8192,0.3478497664133708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,7168,0.33254843287997776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,7168,0.1229706605275472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,6144,0.26028532452053493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,7168,0.3133893277910021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,10240,0.16899911562601724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,8192,0.38031376732720273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,51200,2.4102836185031467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,6144,0.11038488811916775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,5120,0.2743333445654975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,65536,3.113394631279839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,5120,0.09445777866575454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,6144,0.26428177621629506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,5120,0.22310845057169595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,4096,0.2442399925655789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,10240,0.4413119951883952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,3584,0.18895199563768175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,3584,0.06864000029034086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,3584,0.16867733001708984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,3072,0.17040888468424478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,3072,0.14636354976230198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,4096,0.0765742195977105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,6144,0.28531021542019314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,2560,0.11631111303965251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,3072,0.06253688865237765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,2560,0.05475377705362108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,2560,0.12573333581288657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,3584,0.1636542214287652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,2048,0.10398933622572158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,3072,0.14083200030856663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,4096,0.18612444400787354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,1536,0.08664355675379436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,1536,0.08236533403396606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,4096,0.20616800255245635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,1024,0.05110222101211548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,1536,0.03677777780426873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,1024,0.06320177846484713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,2048,0.04531466629770067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,1024,0.06235644552442762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,1024,0.0303422212600708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,5120,0.24052622583177355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,768,0.051903999514049955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,1536,0.08044888575871785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,512,0.03367022342152066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,768,0.05167999863624573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,2048,0.10034755865732829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,512,0.04481777879926893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,768,0.047113776206970215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,512,0.025198222862349615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,512,0.043624000416861645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,2048,0.11926489406161839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,256,0.02330044408639272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,128,0.020555555820465088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,256,0.03480888737572564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,768,0.02769511110252804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,128,0.023081777824295893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,128,0.037787556648254395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,32,0.019176888797018263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,256,0.023172444767422144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,32,0.03735555542839898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,2560,0.11888799402448867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,256,0.04082311193148295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,64,0.018512000640233357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,128,0.028468443287743464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,64,0.03737333416938782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,65536,0.36118578910827637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,51200,0.7801493538750542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,65536,1.1152755949232314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,16384,0.26589687665303546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,51200,0.8245955573187934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,51200,0.29220443301730686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,16384,0.10055822134017944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,12288,0.18102667066786024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,65536,1.0027582380506728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,16384,0.23997065756056044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,12288,0.18771999412112764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,12288,0.07083911365932889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,10240,0.18074222405751547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,12288,0.18179910712771943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,8192,0.14873777495490179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,10240,0.06061689058939616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,10240,0.15306488672892252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,51200,0.7959973547193738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,8192,0.12661333878835043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,7168,0.11117600070105658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,8192,0.12450577153099908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,6144,0.1134328842163086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,7168,0.04531733194986979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,7168,0.11180355813768174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,7168,0.11013866795433892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,6144,0.04034666551484002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,5120,0.08566044436560737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,6144,0.09610578086641099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,5120,0.08193688922458224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,5120,0.03411644366052415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,4096,0.07889066802130805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,4096,0.06796711020999484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,4096,0.06975111034181383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,4096,0.029164443413416546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,6144,0.09752621915605332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,8192,0.049639999866485596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,16384,0.2853439913855659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,3584,0.06248799959818522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,3072,0.05215733249982198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,3584,0.02663733396265242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,3072,0.024099555280473497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,3072,0.052705778015984424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,2560,0.048842665221956044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,3072,0.05539822247293261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,3584,0.06700266732109918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,2048,0.039432889885372586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,2560,0.0214382227924135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,2560,0.049469331900278725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,65536,1.0018364588419597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,1536,0.02846933404604594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,2048,0.01885422236389584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,2048,0.04139822059207492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,1536,0.030385779009924993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,1024,0.02122488949033949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,10240,0.15847467051612005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,2560,0.04561333192719353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,1024,0.023856888214747112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,3584,0.06010755565431383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,1024,0.013800000150998434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,5120,0.08294933372073703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,768,0.01276177747382058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,768,0.01963555481698778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,512,0.013389333254761167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,768,0.023718222975730896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,1536,0.034866667456097074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,512,0.011754666765530905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,512,0.01685422162214915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,512,0.020392000675201416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,256,0.015166223049163818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,1024,0.027753776974148218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,128,0.010372444159454769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,128,0.014714666538768344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,1536,0.01647200021478865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,128,0.011000889043013254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,256,0.012191111014948951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,64,0.010026666853162978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,32,0.010207110808955299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,64,0.014556444353527494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,256,0.016701334052615695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,32,0.014515555567211576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,128,0.01479022204875946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,65536,0.6709458033243815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,65536,0.8053884506225586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,768,0.017664000391960144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,51200,0.5796408653259277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,51200,0.205040004518297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,65536,0.6818524466620551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,16384,0.19160711765289307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,256,0.010802666346232096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,2048,0.03808088766203986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,12288,0.1484702163272434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,16384,0.20622311698065865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,16384,0.06835911009046766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,16384,0.17234310838911268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,10240,0.11078133185704549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,12288,0.1428204509947035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,51200,0.6392595503065321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,12288,0.13401956028408474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,10240,0.12029243840111627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,10240,0.11031199826134576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,65536,0.26560354232788086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,8192,0.0976933307117886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,8192,0.03730577892727322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,8192,0.09178311294979519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,51200,0.5307262208726671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,7168,0.03334577878316244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,7168,0.08662222491370307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,7168,0.08126399914423625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,10240,0.046478221813837685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,8192,0.09364710913764106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,5120,0.0694391131401062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,6144,0.07608799801932441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,6144,0.0708915525012546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,4096,0.05279644330342611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,7168,0.08952799770567153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,5120,0.06469333171844482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,5120,0.06167733007007175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,3584,0.048026667700873486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,4096,0.0533537765343984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,12288,0.05299999978807238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,6144,0.06971822182337443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,3584,0.02023555503951179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,3584,0.045904888047112354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,5120,0.02580533259444767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,3072,0.04172622164090475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,3072,0.018564444449212816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,3072,0.04143911268975999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,4096,0.05169066786766052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,2560,0.03661244445376926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,2560,0.017030222548378837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,2560,0.036650667587916054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,6144,0.029764443635940552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,2048,0.015277332729763456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,2048,0.030453334252039593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,2048,0.03158844510714213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,2560,0.03393422232733832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,4096,0.02217866645918952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,1024,0.015222221612930298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,1536,0.013436444103717804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,1536,0.026608000199000042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,2048,0.02847555610868666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,1024,0.019535111056433786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,768,0.01276177747382058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,3072,0.03426488902833726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,1024,0.021918222308158875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,512,0.010131555298964182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,768,0.010402667025725046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,768,0.016237333416938782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,768,0.0188746667570538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,256,0.007661333514584436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,512,0.013849777479966482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,1536,0.024889777104059856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,256,0.009185777770148383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,128,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,256,0.014161777165200977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,1024,0.011383111278216044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,3584,0.04746133420202467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,64,0.006602666858169768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,128,0.009393778112199571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,32,0.006313777632183499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,128,0.013352000051074557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,64,0.012140444583363004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,512,0.009864889085292816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,512,0.016307555966907077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,256,0.012808000048001608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,65536,0.6021013259887695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,65536,0.24232888221740723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,65536,0.6731306711832682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,51200,0.496155579884847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,1536,0.021291555629836187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,16384,0.17881245083279082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,51200,0.5620382097032335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,51200,0.19675111770629883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,16384,0.16255822446611193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,16384,0.06276088953018188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,12288,0.10681866937213474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,65536,0.657091564602322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,128,0.012247111234400006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,12288,0.050473776128556996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,51200,0.5021742184956869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,10240,0.10764533281326294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,12288,0.12468177742428249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,8192,0.07366755273607042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,32,0.012104888757069906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,10240,0.10331555869844224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,10240,0.10645244518915813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,8192,0.03448977735307481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,16384,0.16701600286695692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,8192,0.0878524449136522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,12288,0.12324978245629205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,7168,0.07464977767732409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,7168,0.031211555004119873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,7168,0.07989955610699125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,10240,0.042075554529825844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,6144,0.06515644656287299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,5120,0.057024889522128634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,6144,0.06875022252400716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,7168,0.0739368862575955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,8192,0.08375910917917888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,5120,0.05939910809199015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,4096,0.021202666891945735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,4096,0.045648889409171216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,3584,0.04336977667278714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,4096,0.049792889091703624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,6144,0.028012444575627644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,3584,0.019469334019554984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,3072,0.038112001286612615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,3584,0.045048889186647206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,4096,0.04306844539112515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,3072,0.03564000129699707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,2560,0.02970222300953335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,3072,0.0409840014245775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,6144,0.0673475530412462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,2560,0.031014220582114324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,2048,0.02522755497031742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,2560,0.03601155678431193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,3584,0.04064977831310696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,1536,0.020777778493033517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,5120,0.055366221401426524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,1536,0.02182577715979682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,2560,0.016720889343155753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,1536,0.025841777523358662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,2048,0.026555554734336004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,2048,0.030733333693610296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,5120,0.024533333049880132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,768,0.012910222013791403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,1536,0.012795555922720166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,1024,0.01610666679011451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,768,0.01444355481200748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,512,0.009696889254781935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,768,0.010065777434243096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,2048,0.014774221513006421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,512,0.012111999922328524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,512,0.009655111365848118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,256,0.0073315559162033936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,512,0.015496888094478183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,1024,0.0214675549003813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,256,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,256,0.011408000356621213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,256,0.013140444126394061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,128,0.011067555182509951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,128,0.00907199995385276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,64,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,32,0.006732444382376141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,64,0.011054221954610614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,3072,0.017993777990341187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,32,0.011063111325105032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,128,0.006593777901596493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,65536,0.4858480029635959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,65536,0.1975688934326172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,65536,0.5503670904371474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,51200,0.37367110782199436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,128,0.012103999654452005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,16384,0.1157955593532986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,1024,0.011023110813564725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,51200,0.4306764337751601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,65536,0.4967164463467068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,16384,0.13469955656263563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,16384,0.05249333381652832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,16384,0.12872444258795843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,51200,0.38939287927415633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,1024,0.01720711092154185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,12288,0.10005422433217366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,10240,0.08647288878758748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,10240,0.08494222164154053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,10240,0.034634666310416326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,10240,0.0915457804997762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,8192,0.07081600030263265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,8192,0.028216001060273912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,7168,0.06781333022647433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,8192,0.0703279972076416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,12288,0.10233510865105523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,7168,0.06313155757056342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,7168,0.06305599874920316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,7168,0.02603111167748769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,6144,0.05287555522388882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,768,0.018594667315483093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,6144,0.023004444109068975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,6144,0.055431998438305326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,5120,0.04099022348721822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,51200,0.1555146641201443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,5120,0.020650666620996263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,5120,0.04711822337574429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,8192,0.07580888933605619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,5120,0.04878488845295376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,4096,0.018227555685573153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,12288,0.10237155357996623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,3584,0.03081511126624213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,3584,0.034565332863065935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,3584,0.037167112032572426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,3072,0.028206222587161597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,4096,0.03824088970820109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,3072,0.03108000093036228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,12288,0.040589332580566406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,4096,0.04086755712827047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,2560,0.024029332730505202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,3072,0.015308444698651632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,2048,0.020224889119466145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,2560,0.014484445254007975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,6144,0.05552266703711616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,2560,0.027055111196306016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,1536,0.01613333324591319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,3584,0.01692800058258904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,2048,0.02590933276547326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,1536,0.019130667050679524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,1024,0.012944888737466602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,4096,0.039289779133266874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,3072,0.03320088982582092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,1024,0.014826666977670459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,1024,0.017927999297777813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,2560,0.029581334855821397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,2048,0.013234666652149625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,768,0.012744888663291931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,512,0.009008000294367472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,768,0.015411555767059326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,1536,0.021877333521842957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,512,0.011062222222487131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,256,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,1024,0.009689778089523315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,768,0.010773333410422007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,256,0.008343111309740279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,128,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,256,0.01184177812602785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,768,0.00906577789121204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,1536,0.011316444310877057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,64,0.006311111152172089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,128,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,32,0.006258666515350342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,128,0.01132266637351778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,512,0.013518222504191928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,64,0.009699555734793345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,32,0.009675555759006077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,65536,0.45424355400933164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,65536,0.49266576766967773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,128,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,512,0.008691555923885768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,65536,0.4892435603671604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,51200,0.14208355214860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,2048,0.02295822236273024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,16384,0.11474843819936116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,51200,0.3810266653696696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,65536,0.18671911292605928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,16384,0.12047910690307617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,12288,0.08662488725450303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,12288,0.038204444779290095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,12288,0.09220444493823582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,10240,0.09203910827636719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,51200,0.3910240067376031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,256,0.010079111489984725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,10240,0.07845866680145264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,8192,0.07820266485214233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,10240,0.08374666505389744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,51200,0.37458843655056423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,16384,0.12609955999586317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,8192,0.027393778165181477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,8192,0.07078044282065497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,7168,0.025070221887694463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,7168,0.057355556223127574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,6144,0.05858933263354831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,10240,0.03321689036157396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,6144,0.049633777803844877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,6144,0.022240888741281297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,6144,0.05572799841562907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,8192,0.06429777542750041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,7168,0.06673955255084567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,5120,0.01996977792845832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,5120,0.04861866765552097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,12288,0.09806311130523682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,7168,0.06200889084074232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,4096,0.0359173317750295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,4096,0.04121422105365329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,5120,0.04828622274928623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,3584,0.03235110971662734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,3584,0.016303999556435477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,3072,0.02960088849067688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,4096,0.0369395547442966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,3072,0.028584000137117174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,4096,0.01758577757411533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,3072,0.03272355596224467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,2560,0.02532888948917389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,16384,0.04881244566705492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,2560,0.025246222813924152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,2560,0.013935110635227628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,2560,0.029363556040657893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,3584,0.0374017788304223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,5120,0.04249155521392822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,1536,0.017232000827789307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,2048,0.021519111262427434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,2048,0.025449777642885845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,3584,0.03397688931889004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,1024,0.011402666568756104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,3072,0.014986667368147107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,1536,0.01090400003724628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,1536,0.02160533269246419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,768,0.009376000199053023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,1024,0.013817778064144982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,2048,0.02091288897726271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,768,0.012087110843923358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,2048,0.012698666916953193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,768,0.015124445160230001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,512,0.008628444539176093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,1536,0.017601778109868366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,256,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,512,0.013129777378506131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,1024,0.009681777821646797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,256,0.008064000142945183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,128,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,256,0.011304000185595619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,512,0.0074026667409473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,128,0.009324444664849175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,64,0.0052382221652401825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,128,0.010943111446168689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,512,0.010453333457310995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,64,0.009006222089131674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,32,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,65536,0.34160534540812176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,256,0.009495110975371467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,1024,0.01720622181892395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,65536,0.4217395517561171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,128,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,51200,0.3150435553656684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,768,0.009001777403884465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,32,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,16384,0.11061066389083862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,51200,0.11414489481184219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,51200,0.33659201198154026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,16384,0.03853066762288412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,16384,0.1046631137530009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,16384,0.09622222185134888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,51200,0.2802080048455132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,10240,0.06604088677300347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,12288,0.07412177986568876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,65536,0.141994661755032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,10240,0.06330311298370361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,12288,0.08044089211357965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,8192,0.05513599846098158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,8192,0.02143999934196472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,7168,0.047841777404149376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,8192,0.0525493323802948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,10240,0.07112177875306872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,8192,0.05169777737723457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,6144,0.03968711031807794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,7168,0.04755199948946635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,65536,0.3574568960401747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,6144,0.043022221989101835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,6144,0.042039109600914844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,5120,0.03676977753639221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,7168,0.049356444014443286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,7168,0.01994400057527754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,4096,0.02793866727087233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,5120,0.0365920000606113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,6144,0.018056000272432964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,4096,0.031209776798884075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,3584,0.024800888366169397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,4096,0.014640889234013028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,4096,0.03161333335770501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,10240,0.02611911131276025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,3584,0.028121779362360638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,3584,0.028581334484948054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,5120,0.0162791113058726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,3072,0.025240000751283433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,3072,0.012736000120639801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,3072,0.026128000683254663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,12288,0.07961244715584649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,2560,0.021935110290845234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,2048,0.015584000282817416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,2560,0.02351555559370253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,3584,0.013542222479979197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,3072,0.020872889293564692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,1536,0.01309866706530253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,2048,0.01036622209681405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,2048,0.019056888090239633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,2048,0.020850666695170935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,1536,0.009336888790130615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,2560,0.018546667363908555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,1024,0.010040889183680216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,12288,0.030244443151685927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,1024,0.01220533334546619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,768,0.008309333688682979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,768,0.0107351111041175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,1536,0.015487111277050443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,5120,0.03289955523278978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,768,0.012882666455374824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,2560,0.011447110937701331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,1024,0.014754666222466363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,512,0.009374221993817223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,256,0.005618666609128316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,512,0.011424889167149862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,256,0.00869777798652649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,256,0.010040000081062317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,768,0.00811644477976693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,1024,0.008391111261314815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,128,0.008665777742862701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,512,0.007882666256692674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,128,0.010456889039940305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,32,0.004484444442722532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,64,0.00868444475862715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,256,0.007370666497283512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,512,0.00674222202764617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,65536,0.3672942320505778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,128,0.007369777394665613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,65536,0.134807997279697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,65536,0.36220534642537433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,51200,0.2153244415918986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,65536,0.3394426769680447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,51200,0.28357601165771484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,32,0.008745777938101027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,51200,0.26623643769158256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,16384,0.09122221999698216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,16384,0.03742311067051358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,64,0.004527111020353106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,12288,0.0711359977722168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,16384,0.09231555461883545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,10240,0.057800889015197754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,12288,0.029100444581773546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,12288,0.07044711377885607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,12288,0.07207110855314466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,1536,0.017261332935757108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,10240,0.02592977715863122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,51200,0.10465066962771946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,10240,0.06166399849785698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,7168,0.0423351095782386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,8192,0.0210853334930208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,8192,0.04901955525080363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,8192,0.052806221776538424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,6144,0.034597333934572004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,7168,0.04340799980693393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,7168,0.04695733388264974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,6144,0.038482666015625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,6144,0.04120533333884345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,10240,0.05950399902131823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,5120,0.03288622366057502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,5120,0.035967111587524414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,7168,0.01953155630164676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,5120,0.016117332710160148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,4096,0.0281368891398112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,4096,0.031158222092522517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,128,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,3584,0.022385777698622808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,3584,0.02512977851761712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,3584,0.027950220637851294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,3072,0.019955555597941082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,3584,0.013166222307417126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,3072,0.02275644408331977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,5120,0.029303110308117334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,3072,0.025455110602908667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,2560,0.017529777354664274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,2560,0.019934222102165222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,2560,0.023366222778956097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,2560,0.011319110790888468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,2048,0.017095110482639737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,6144,0.017948443690935772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,4096,0.01419911119672987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,2048,0.009700444837411245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,2048,0.02040177749262916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,1536,0.013854222165213691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,4096,0.02444355520937178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,1024,0.009728888670603434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,1536,0.00922400007645289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,16384,0.0764240026473999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,1024,0.01145155562294854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,3072,0.012367111113336353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,1024,0.014262222581439547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,1536,0.012413333687517377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,768,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,768,0.010030222435792288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,768,0.012776888906955719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,2048,0.0151573336786694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,1536,0.01683911184469859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,512,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,512,0.011744000017642975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,8192,0.04573422339227465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,768,0.00796888851457172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,256,0.0073724447025193115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,128,0.005159111072619756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,256,0.01036888857682546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,512,0.006948444578382704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,128,0.00793333351612091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,64,0.004580444345871608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,256,0.005648889061477449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,32,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,64,0.00796888851457172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,256,0.008355555435021719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,1024,0.008395555946562026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,65536,0.29860978656344944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,128,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,128,0.010377777947319878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,512,0.008839999636014303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,51200,0.23889867464701334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,65536,0.3380631075965033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,16384,0.06299466556972928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,51200,0.09957422150505914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,51200,0.24571198887295195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,51200,0.2672942214541965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,16384,0.03529777791765001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,12288,0.059624003039466016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,16384,0.0913848876953125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,65536,0.3121448887719048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,10240,0.049597332874933876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,12288,0.06129066811667549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,12288,0.070997330877516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,10240,0.024822221861945257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,10240,0.06119910875956217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,32,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,8192,0.04281244344181485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,8192,0.020557334025700886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,7168,0.031339555978775024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,8192,0.05082133412361145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,7168,0.03853333327505324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,12288,0.028282665544086035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,6144,0.03145066565937466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,7168,0.04587822159131368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,10240,0.0525493323802948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,5120,0.024878222081396315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,6144,0.0340320004357232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,6144,0.04023911224471198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,5120,0.015506666567590503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,4096,0.021233777205149334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,5120,0.035600887404547796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,65536,0.12845155927870008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,7168,0.019147555033365887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,4096,0.024959110551410254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,4096,0.030605332718955144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,8192,0.04016266597641839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,6144,0.017482666505707633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,3072,0.016844444804721408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,3584,0.022670222653283015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,3584,0.02777155571513706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,5120,0.029488000604841445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,2560,0.014993778533405729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,3072,0.011762667033407422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,3072,0.025182222326596577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,4096,0.013842666314707862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,2560,0.01109155515829722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,3584,0.018935110833909776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,2560,0.02216622233390808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,2048,0.01036444471942054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,1536,0.010344888601038191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,2048,0.019935111204783123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,16384,0.07953421937094794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,3072,0.020426667398876615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,1024,0.007631999750932057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,1536,0.00907199995385276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,1536,0.016587555408477783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,2560,0.017839110559887357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,2048,0.013125333521101209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,1024,0.008447111480765873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,1024,0.013826666606797112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,2048,0.01517955627706316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,768,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,768,0.012465777496496836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,512,0.008344000412358178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,512,0.011335111326641507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,256,0.004946666873163647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,768,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,3584,0.012783110969596438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,768,0.009022222624884712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,128,0.004569777597983678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,256,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,256,0.010031999813185798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,1536,0.012551999754375882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,64,0.004217777815130022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,128,0.0075288886825243635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,128,0.010072000324726105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,32,0.004234666625658671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,1024,0.010375111467308467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,64,0.007335999773608313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,32,0.007285333342022366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,512,0.0058871110280354815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,256,0.00776977754301495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,51200,0.1644888851377699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,128,0.007326222128338284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,512,0.007660444411966536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,51200,0.21350933445824516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,51200,0.09688978062735663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,65536,0.22554400232103136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,51200,0.26328354411655003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,16384,0.034268445438808866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,16384,0.07078311178419325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,12288,0.05304888884226481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,65536,0.27878933482699925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,16384,0.09194844298892552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,10240,0.04161422120200263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,12288,0.02737066646416982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,65536,0.33849512206183535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,12288,0.05400800042682224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,12288,0.07000088691711426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,10240,0.024453333682484094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,10240,0.046318223079045616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,16384,0.0678222245640225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,10240,0.06030488676495022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,8192,0.038169778055614896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,8192,0.050254222419526845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,65536,0.12238932980431451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,7168,0.0344471103615231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,7168,0.04524088899294535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,6144,0.030403554439544678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,6144,0.03970488905906677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,7168,0.03046755658255683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,8192,0.020629333125220407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,5120,0.015448888142903646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,6144,0.027119111683633592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,4096,0.02015022271209293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,5120,0.026344888740115698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,4096,0.022505778405401442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,4096,0.013343111508422427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,3584,0.016908443636364408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,6144,0.016864889197879367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,5120,0.021531555387708876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,3072,0.014239999983045789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,3584,0.027493332823117573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,7168,0.01868266695075565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,3072,0.01791822248035007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,3072,0.024719999896155462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,8192,0.035134222772386335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,4096,0.030247999562157526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,2560,0.016327111257447135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,2048,0.011544889046086205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,3584,0.01274577776590983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,2048,0.013629333012633853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,2048,0.019288889235920377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,2560,0.013103110922707452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,5120,0.03567733367284139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,1536,0.01144444445768992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,2560,0.010556444525718689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,1024,0.007343110938866933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,3584,0.020382222202089097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,1024,0.009894222021102905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,3072,0.011424000064531961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,768,0.006607999818192587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,1024,0.008377778033415476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,768,0.00867022242810991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,2560,0.022250667214393616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,768,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,2048,0.00962311112218433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,1536,0.009678222239017487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,512,0.007631999750932057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,256,0.004567111117972268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,512,0.011063999599880643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,1536,0.01609422266483307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,256,0.007295110987292395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,128,0.004212444441186057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,256,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,1024,0.013809777796268463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,128,0.006959999601046245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,64,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,128,0.01017955525053872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,32,0.003896000070704354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,64,0.007136888802051544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,1536,0.008904889225959778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,512,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,32,0.006990222467316522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,65536,0.16892977555592856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,51200,0.1372284491856893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,512,0.007727111379305522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,65536,0.2396622233920627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,65536,0.19490310880872938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,128,0.007429333196745978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,51200,0.1891120009952121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,51200,0.1540693309572008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,768,0.012151999606026543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,12288,0.040243556102116905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,256,0.01036266651418474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,16384,0.05634311172697279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,65536,0.08140977885988024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,12288,0.019536889261669584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,10240,0.03609600000911289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,12288,0.04403555393218994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,16384,0.057188444667392306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,10240,0.017515555024147034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,8192,0.02879022227393256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,10240,0.03939466675122579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,16384,0.06080089012781779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,8192,0.03307999836073981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,7168,0.0255039996571011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,12288,0.046634667449527316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,7168,0.030108445220523413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,10240,0.040022220876481794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,7168,0.030413332912656996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,16384,0.023775999744733173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,6144,0.02663999961482154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,6144,0.012470222181744046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,5120,0.019285332825448778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,6144,0.02699022160636054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,51200,0.06709244516160753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,5120,0.023212444451120164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,6144,0.022479999396536086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,4096,0.016542222764756944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,8192,0.03314133485158285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,4096,0.019671999745898776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,3584,0.014839111102951897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,4096,0.010046222143703038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,3584,0.018173333671357896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,3072,0.013487999637921652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,7168,0.013598221871587964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,5120,0.023570666710535686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,3072,0.015948444604873657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,2560,0.012075555821259817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,3072,0.017218665944205392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,4096,0.021143111917707656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,2560,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,2048,0.010391111175219217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,2560,0.015511999527613321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,3584,0.019356444478034973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,5120,0.011065777805116443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,2048,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,2048,0.012321777641773224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,2048,0.014154665999942355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,3072,0.009351111120647853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,1024,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,1536,0.010384000009960597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,1536,0.012497777740160624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,2560,0.014076444837782117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,768,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,3584,0.009716444545321995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,768,0.0079297779334916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,8192,0.014831110835075378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,512,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,1536,0.007974222302436829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,1024,0.008653333617581261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,512,0.00720355576939053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,1024,0.010936000280910067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,512,0.009362666971153682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,512,0.0069475554757648045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,256,0.006663110935025745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,1536,0.008314666648705801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,256,0.008676444490750631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,768,0.010093332992659675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,128,0.0069511110583941145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,128,0.006726222319735422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,64,0.003571555432346132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,128,0.008742222355471717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,1024,0.0073902226156658595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,256,0.004238222208287981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,64,0.006844444407357111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,32,0.006612444503439798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,256,0.006682666639486949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,65536,0.07799822092056274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,768,0.007167110840479533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,65536,0.2235342131720649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,32,0.0035422220826148987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,65536,0.18916800287034777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,51200,0.17644266287485758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,51200,0.14815910657246908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,65536,0.16069511572519937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,12288,0.03494222296608819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,16384,0.023137778043746948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,128,0.003843555433882607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,51200,0.16110222869449192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,12288,0.04505066739188301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,12288,0.04360888732804192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,51200,0.06128444274266561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,10240,0.03872266742918227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,8192,0.02548622257179684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,10240,0.03779110974735684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,10240,0.01718488832314809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,16384,0.05855911307864719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,16384,0.054993778467178345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,8192,0.03247377938694424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,8192,0.03237688872549269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,10240,0.032256000571780734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,6144,0.020076443751653034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,7168,0.013367111484209696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,7168,0.029731555117501154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,16384,0.04831644561555651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,5120,0.016991110311614144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,6144,0.026185777452256944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,8192,0.014398222168286642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,12288,0.019347555107540555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,5120,0.011090666883521609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,4096,0.014305777019924588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,7168,0.02933688958485921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,4096,0.010060444474220276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,6144,0.012252444194422828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,3584,0.013408888545301227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,7168,0.022219555245505437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,3584,0.017052445146772597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,5120,0.023415999280081853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,3072,0.011888888974984487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,3584,0.018385777870814007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,4096,0.019055111540688407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,3072,0.015649777319696214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,6144,0.02606933315594991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,3072,0.009091555244392818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,3072,0.016992888516849942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,5120,0.022557333111763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,2048,0.009012444151772393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,2560,0.008825778133339351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,2048,0.012116444607575735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,2048,0.008358221915033128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,2560,0.01073066641887029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,2048,0.013853333062595792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,2560,0.01367911116944419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,1536,0.012139555480745105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,1536,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,3584,0.009692444569534725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,2560,0.015536889433860779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,1024,0.0073724447025193115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,768,0.005470222069157495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,1024,0.011062222222487131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,4096,0.02068800065252516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,512,0.004576000074545543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,768,0.007032000356250339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,768,0.007701333198282454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,768,0.010053333308961656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,256,0.004230222354332606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,512,0.006888888776302338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,1024,0.005625777774386936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,512,0.009533333281675974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,128,0.003887999802827835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,256,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,256,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,1024,0.008723555339707268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,128,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,64,0.0037377778854635027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,32,0.003912000192536248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,128,0.008665777742862701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,65536,0.1626782284842597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,1536,0.010060444474220276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,32,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,65536,0.07338666915893555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,51200,0.13091911209954157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,65536,0.18433954980638292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,128,0.006367110957702001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,1536,0.007328888608349695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,16384,0.04313244422276815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,51200,0.05930844280454847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,64,0.006314666734801398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,512,0.006952889263629913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,12288,0.03149155444569058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,51200,0.146433777279324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,256,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,12288,0.018570666511853535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,10240,0.026131555438041687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,12288,0.04255200094646878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,51200,0.1451644500096639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,10240,0.03305066625277201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,10240,0.037104888094796076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,10240,0.016910221841600206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,16384,0.05452533231841194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,65536,0.18802755408816865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,7168,0.019186667270130582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,8192,0.027487110760476854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,8192,0.013832000394662222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,12288,0.03784177700678507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,8192,0.031903998719321355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,6144,0.016505777835845947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,7168,0.013021333350075616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,7168,0.0247999992635515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,5120,0.014808000789748298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,6144,0.02235377828280131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,6144,0.025669333007600572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,8192,0.02134311033619775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,16384,0.022152889106008742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,5120,0.011050666371981302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,5120,0.023395554886923894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,4096,0.015818667080667283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,4096,0.009880888793203566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,4096,0.020204444726308186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,7168,0.02886577778392368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,3584,0.014760888285107084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,3584,0.018004443910386827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,16384,0.04877510997984144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,3072,0.01346577786737018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,3072,0.00903733323017756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,2560,0.009381333159075843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,6144,0.01186222251918581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,2560,0.012098666694429187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,3072,0.010464888479974536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,2048,0.008218666745556725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,5120,0.019147555033365887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,2560,0.015288889408111572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,2048,0.00832444429397583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,2048,0.010709332923094431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,3072,0.016215110818545025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,1536,0.0069520001610120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,1536,0.009024000002278222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,3584,0.009587555295891231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,1024,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,4096,0.012808000048001608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,1024,0.007706666986147563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,1024,0.011066666907734342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,768,0.005249777601824866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,768,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,1536,0.012129777835475074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,1536,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,2560,0.00868800034125646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,768,0.007038222418891058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,512,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,3584,0.011426666544543372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,1024,0.007481777833567725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,512,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,512,0.009387555221716562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,128,0.003908444609906939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,256,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,256,0.00868444475862715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,512,0.004571555389298333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,64,0.0038764443662431505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,128,0.005978666659858491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,32,0.003567111161020067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,2048,0.013581333061059317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,256,0.004222222086456087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,32,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,256,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,65536,0.1452817784415351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,51200,0.10982755819956462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,65536,0.17619644271002877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,128,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,128,0.008400888906584846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,16384,0.03231199913554721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,51200,0.05786133474773831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,64,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,65536,0.13348978095584446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,16384,0.021665778425004747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,12288,0.02454400062561035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,16384,0.050966223080952965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,12288,0.03046755658255683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,12288,0.017824000782436795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,10240,0.021073778470357258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,12288,0.04004977809058295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,10240,0.016174222032229107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,10240,0.026594666971100703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,768,0.009952888720565373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,51200,0.11332355605231391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,51200,0.1389697790145874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,8192,0.013172444370057849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,16384,0.039001779423819646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,8192,0.022361778550677832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,8192,0.03013066781891717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,7168,0.012316444681750404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,7168,0.020067556036843192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,7168,0.027028444740507338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,6144,0.01811377704143524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,5120,0.01238666640387641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,6144,0.024381332927280005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,10240,0.03478399912516276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,5120,0.010637332995732626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,7168,0.01577333278126187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,4096,0.011067555182509951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,5120,0.021735999319288466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,6144,0.013770666387346057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,4096,0.009967111051082611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,3584,0.010423999693658618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,65536,0.07379466957516141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,4096,0.018575110369258456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,3072,0.008971555365456475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,6144,0.01145866678820716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,3584,0.012478222449620565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,8192,0.017870222528775532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,3584,0.01737600068251292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,3072,0.009020444419648912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,3072,0.015564444992277356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,2560,0.01036977767944336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,2048,0.006656888872385025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,2560,0.014482667048772177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,4096,0.013392888837390475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,2048,0.009452444811662039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,3584,0.009394666386975182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,1536,0.006019555446174409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,2560,0.007713778151406183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,3072,0.011415999796655444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,1024,0.004903111192915174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,1536,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,5120,0.015612444943851896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,1536,0.011400889191362592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,2560,0.008664000365469191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,768,0.004561777744028303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,1024,0.010458666417333813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,768,0.006605333338181178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,768,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,512,0.0038986665507157645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,2048,0.008062221937709385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,768,0.009825777676370408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,512,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,256,0.0035413333939181436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,512,0.008663111262851292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,1024,0.00701422244310379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,1024,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,256,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,256,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,2048,0.013207111093733044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,128,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,64,0.0032551110618644287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,1536,0.008035555481910706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,32,0.0032400000426504347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,64,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,256,0.005608888963858287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,65536,0.08015288909276326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,128,0.003552000141806073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,65536,0.07094755437639025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,65536,0.13850043879614934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,51200,0.07360533210966322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,65536,0.1740657753414578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,51200,0.05708711014853584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,51200,0.10831999778747559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,16384,0.0258577780591117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,51200,0.1380995512008667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,12288,0.02134311033619775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,16384,0.02106933295726776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,16384,0.04987555411126879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,128,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,12288,0.029575109481811523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,512,0.006744888921578725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,12288,0.039635555611716375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,10240,0.01584533353646596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,8192,0.01516355574131012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,32,0.00564444437623024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,10240,0.034112887250052557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,16384,0.0374035570356581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,8192,0.021551110678248938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,7168,0.013888888888888888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,8192,0.012929778132173749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,8192,0.02938400043381585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,6144,0.012791111237472959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,10240,0.01832355558872223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,128,0.008348444269763099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,7168,0.026346666945351496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,5120,0.010810666614108615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,6144,0.017324444320466783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,6144,0.023239110906918842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,10240,0.025663110944959853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,4096,0.009008000294367472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,5120,0.01517599986659156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,5120,0.020727111233605277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,4096,0.013140444126394061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,3584,0.008222222328186035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,7168,0.012133333418104382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,6144,0.01108000013563368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,3584,0.009135111338562435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,3584,0.012289777398109436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,5120,0.01038044442733129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,3584,0.016993777619467843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,4096,0.009676444033781687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,3072,0.015488889482286243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,12288,0.01752622259987725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,7168,0.019479110836982727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,2560,0.008340444829728868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,2560,0.010040000081062317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,2560,0.014150222142537435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,3072,0.007603555917739868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,2048,0.009072889056470659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,2048,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,2048,0.012817777693271637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,3072,0.008821333448092142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,3072,0.011145778000354767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,1536,0.007903999752468532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,1024,0.004575110971927643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,2048,0.006070222291681502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,1024,0.0069884442620807225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,1024,0.00720000018676122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,768,0.00407644444041782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,1024,0.010382222632567087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,768,0.006664888726340399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,768,0.009357333183288574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,512,0.0038933331767717996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,512,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,1536,0.00775999989774492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,512,0.009158222211731805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,512,0.006647999915811751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,1536,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,256,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,128,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,256,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,768,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,2560,0.0069013333155049225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,64,0.003257777748836411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,128,0.006270222365856171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,128,0.005624889085690181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,128,0.008026666939258575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,1536,0.011063111325105032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,768,65536,0.06725777520073785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,32,0.005566222386227713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,65536,0.09655111365848118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,65536,0.07179911269081964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,65536,0.17289066314697266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,768,51200,0.05406755540106031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,4096,0.01791200041770935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,32,0.0035342222286595237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,51200,0.0565591123369005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,51200,0.07630755503972371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,256,0.0035751110149754416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,16384,0.027730666928821143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,16384,0.021218667427698772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,16384,0.04941777719391716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,256,0.006320000108745363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,12288,0.0221395558781094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,768,10240,0.016778666112158034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,12288,0.01719999975628323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,768,16384,0.020945777495702107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,64,0.005633777628342311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,10240,0.019215111931165058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,10240,0.015359110302395292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,10240,0.033879998657438494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,768,12288,0.019375110665957134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,768,7168,0.013212444053755866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,8192,0.028855111863878038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,12288,0.03905066516664293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,768,6144,0.011669333610269757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,7168,0.026551110876931086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,51200,0.1369253396987915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,768,8192,0.014287999934620328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,768,5120,0.010077333284748925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,6144,0.013791110780504016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,8192,0.0127688886390792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,8192,0.01623911162217458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,768,4096,0.00868711123863856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,5120,0.0121760004096561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,5120,0.020232889387342665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,7168,0.014849777022997538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,4096,0.009483555952707926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,768,3584,0.008018666671382057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,4096,0.01793955597612593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,3584,0.009349333743254343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,3584,0.010017777482668558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,768,3072,0.007470221983061896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,3584,0.016248888439602323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,6144,0.0229066660006841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,768,2560,0.006601777755551868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,3072,0.008685333033402761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,5120,0.010217777556843227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,3072,0.015218666858143277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,2560,0.008592888712882996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,2560,0.00869066682126787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,4096,0.01057866629627016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,7168,0.012087999946541257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,768,1536,0.005024888863166173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,2048,0.007811555431948767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,2048,0.01275288893116845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,6144,0.011111111276679568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,768,1024,0.004294222427739037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,1536,0.011111111276679568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,3072,0.009145778086450364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,1024,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,768,768,0.0039057777159743835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,2560,0.013814222481515674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,2048,0.008041777544551426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,768,0.006060444646411472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,768,0.00904088881280687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,768,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,1536,0.007627555893527136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,512,0.006641777853171031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,768,256,0.0032239999208185407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,768,2048,0.006007110906971826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,1024,0.010409778190983666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,256,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,768,128,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,256,0.008371555142932469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,1536,0.006835555450783835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,128,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,512,0.005630222045713001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,1024,0.007111999723646376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,768,32,0.002995555599530538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,64,0.005041777673694822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,256,0.006324444380071428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,512,65536,0.048262221945656665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,32,0.0052266667286554975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,768,512,0.0036595554815398324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,512,51200,0.03871022330390082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,65536,0.09605244133207534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,65536,0.17281244860755074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,768,64,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,512,0.00906044410334693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,51200,0.07466755310694377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,51200,0.055735111236572266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,51200,0.13661332925160727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,512,12288,0.016207999653286405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,16384,0.019813333948453266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,16384,0.02715733316209581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,12288,0.021418665846188862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,65536,0.07067910830179851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,512,10240,0.01497866710027059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,12288,0.0388373335202535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,12288,0.016559110747443307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,128,0.008676444490750631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,10240,0.018606222338146634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,10240,0.03279822071393331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,8192,0.015992888145976596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,8192,0.028121779362360638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,8192,0.012598222328556908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,128,0.005223999834722943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,16384,0.04957155717743767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,7168,0.014541332920392355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,7168,0.012089778151777057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,512,16384,0.019354666272799175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,7168,0.025398222936524287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,6144,0.011076444553004371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,512,5120,0.009376889301670922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,6144,0.022562666071785822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,6144,0.013495999905798169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,512,7168,0.012187555432319641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,512,4096,0.008024888734022776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,5120,0.01014311114947001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,5120,0.019996444384257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,10240,0.014984000060293408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,512,6144,0.010379555324713389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,512,8192,0.013550221920013428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,3584,0.009754666851626502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,3584,0.0162764439980189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,512,3072,0.0070062221752272705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,5120,0.011867555479208628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,3072,0.009073778159088558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,3072,0.008919999831252629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,512,3584,0.007631999750932057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,3072,0.014871110518773397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,2560,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,2560,0.008387555678685507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,512,2048,0.005710222240951326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,2560,0.0138568886452251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,3584,0.009038222332795462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,512,1536,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,2048,0.007672888537247975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,2048,0.012440889245933957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,1536,0.006695111178689533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,512,2560,0.006320000108745363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,4096,0.009341333475377824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,1536,0.011328000161382886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,1024,0.006259555617968242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,4096,0.017901332841979135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,1024,0.007048000064161088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,2048,0.008023111356629265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,768,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,512,512,0.0035297779573334586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,768,0.006029333505365584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,1536,0.007650666766696506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,4096,0.01038755559258991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,768,0.009480888644854227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,512,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,512,0.00545155546731419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,256,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,512,128,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,256,0.008018666671382057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,256,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,128,0.006335111127959357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,512,64,0.0029324444217814338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,128,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,512,32,0.0029111111329661477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,512,256,0.0032142222755485107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,256,65536,0.032093332873450384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,64,0.005004444470008214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,32,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,1024,0.010098666780524785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,256,51200,0.028221332364612158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,65536,0.09454666905932957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,65536,0.07022488779491849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,65536,0.17266489399804008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,128,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,256,16384,0.013807111316257052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,51200,0.07321600119272868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,512,1024,0.004238222208287981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,256,12288,0.011287111375066968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,16384,0.02641511129008399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,16384,0.04872711168395149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,512,768,0.004231111043029361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,12288,0.020967110991477966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,12288,0.016565332810084026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,512,0.008797333472304875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,10240,0.01829866733815935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,10240,0.015222221612930298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,256,8192,0.009339555270142024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,51200,0.13636088371276855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,10240,0.032519999477598406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,256,7168,0.009472000102202097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,8192,0.01275466630856196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,8192,0.02780533168050978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,8192,0.015947555502255756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,256,6144,0.008748444418112436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,7168,0.011749332977665795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,12288,0.037448889679378934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,7168,0.025236444340811834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,7168,0.014639111028777229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,6144,0.010718222293588849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,51200,0.054824001259273954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,6144,0.022011554903454248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,16384,0.01961688862906562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,5120,0.011795555551846823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,256,10240,0.010747555229398938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,4096,0.010413332945770687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,256,5120,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,256,3584,0.007296889192528195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,4096,0.017584888471497428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,5120,0.01016355554262797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,3584,0.009738666315873465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,256,3072,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,3584,0.009019555317031013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,3584,0.016563556260532804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,3072,0.008719999757077958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,3072,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,3072,0.014843554960356818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,4096,0.009400889277458191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,2560,0.008316444853941599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,256,2048,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,6144,0.01329066687160068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,5120,0.019965334071053397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,2048,0.008008889026112026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,256,1536,0.004906666775544484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,2048,0.012439111040698158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,256,2560,0.006206222292449739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,256,1024,0.003933333274390963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,1536,0.006687110910813014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,1536,0.011037333144081963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,2560,0.008372444245550368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,2560,0.013795555465751223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,256,768,0.003924444317817688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,256,4096,0.007640889121426477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,768,0.00675022229552269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,768,0.005894222193294101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,768,0.009038222332795462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,1536,0.0074879998962084455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,512,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,256,256,0.0031928889867332247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,512,0.006704888823959563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,1024,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,512,0.008774222599135505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,1024,0.009751111268997192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,256,128,0.0032879999942249725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,256,0.006350222147173352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,2048,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,128,0.0063786668082078295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,128,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,256,32,0.0028826666788922418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,128,0.008014221986134848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,64,0.005048000150256687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,128,65536,0.026467555099063452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,32,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,1024,0.0069955554273393415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,128,51200,0.018591110905011494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,65536,0.17293955220116508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,256,512,0.003546666767862108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,256,0.008086221913496653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,128,16384,0.012248888611793518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,51200,0.13686133755577937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,16384,0.02583733366595374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,256,64,0.0028444443725877335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,16384,0.04798310995101929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,51200,0.0541928874121772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,12288,0.020996444755130347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,12288,0.03735911183887058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,65536,0.0928657783402337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,16384,0.01978488928741879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,10240,0.01829333272245195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,10240,0.015192889504962497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,128,8192,0.008752889103359645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,10240,0.03253333436118232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,12288,0.016569778323173523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,65536,0.06987022029028998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,128,7168,0.008347555167145198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,8192,0.027641778190930683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,51200,0.07304266426298353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,7168,0.014188443620999655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,7168,0.011753777662913004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,7168,0.02477600011560652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,6144,0.013083555632167392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,256,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,6144,0.022327999273935955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,8192,0.01276000009642707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,128,12288,0.01038577738735411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,5120,0.011900444825490316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,128,6144,0.008015111088752747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,128,4096,0.007744000189834171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,5120,0.020101333657900494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,4096,0.010096888575288985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,128,3584,0.007328888608349695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,128,5120,0.007646222081449296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,4096,0.01754844519827101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,3584,0.009731555150614845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,3584,0.009030222064918941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,128,3072,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,3584,0.01607466737429301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,5120,0.010105777945783403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,128,2560,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,3072,0.00869422240389718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,3072,0.008990222381220924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,3072,0.014481777946154276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,2560,0.008349333372380998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,8192,0.015615999698638916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,2560,0.013794666363133324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,2048,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,2048,0.012425777812798819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,2048,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,1536,0.006688888702127669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,2560,0.00832177781396442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,128,10240,0.009694221946928237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,6144,0.010757333702511258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,1536,0.007350222104125553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,1536,0.01107377807299296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,1024,0.005965333431959152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,128,768,0.003552888830502828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,128,1536,0.004571555389298333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,768,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,768,0.008998221821255153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,128,2048,0.0052355556852287715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,768,0.006653333289755716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,128,1024,0.003929777691761653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,512,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,512,0.005437333136796951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,512,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,1024,0.010028444230556488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,1024,0.007033777733643849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,128,128,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,256,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,256,0.00812533332241906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,128,64,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,128,0.004975999808973736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,128,32,0.0028844444702068963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,4096,0.0094355551732911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,64,65536,0.02589599953757392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,64,0.005015999906592899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,32,0.004975111120276981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,65536,0.09255111217498779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,64,16384,0.009407999614874521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,51200,0.07315199904971652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,64,12288,0.008387555678685507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,128,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,64,10240,0.00906844437122345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,128,0.007991111112965478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,16384,0.025478222303920325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,12288,0.0206675562593672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,64,7168,0.00830666638082928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,10240,0.01789688898457421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,64,51200,0.020394666327370536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,64,6144,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,64,5120,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,8192,0.01555199921131134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,64,4096,0.007734222544564142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,6144,0.013155555559529198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,5120,0.011459555890825061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,64,3584,0.00739644467830658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,64,3072,0.007087111473083496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,3584,0.009715555442704095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,64,2560,0.006393777827421824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,3072,0.008912000391218398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,64,2048,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,7168,0.014182221558358936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,2560,0.008047999607192146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,2048,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,128,512,0.0031964443624019623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,64,8192,0.008385777473449707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,64,768,0.003539555602603488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,128,256,0.0028782222006056043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,64,512,0.003206222214632564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,768,0.005697777701748743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,64,256,0.0029191111938820947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,256,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,512,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,64,1536,0.004267555558019214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,64,64,0.0025822222232818604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,64,1024,0.00388355553150177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,64,32,0.002543111021320025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,4096,0.010219555762079027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,1536,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,64,0.005015999906592899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,1024,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,32,0.00499555551343494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,32,16384,0.014324444863531323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,65536,0.09344266520606147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,32,12288,0.009375111096435124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,51200,0.07277777459886339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,64,128,0.0026542221506436667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,32,10240,0.008718222379684448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,32,8192,0.00868000007337994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,16384,0.025423111187087163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,32,7168,0.00833777752187517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,128,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,10240,0.01793155570824941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,32,6144,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,32,65536,0.01938222183121575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,8192,0.01552088889810774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,32,4096,0.007654222349325816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,7168,0.014151111245155334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,32,3584,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,5120,0.01143555591503779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,32,3072,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,3584,0.009524444739023844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,256,0.005096000101831225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,3072,0.009003555609120263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,32,2560,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,32,2048,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,32,5120,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,32,1536,0.004569777597983678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,32,51200,0.012447111308574677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,2048,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,4096,0.010087110930018956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,1536,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,12288,0.020464888877338834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,32,512,0.0032293332947625052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,768,0.005707555760939916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,32,256,0.0031999999450312722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,2560,0.0083155557513237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,32,128,0.002549333290921317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,32,1024,0.0038844446341196695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,32,64,0.0025600000388092464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,256,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,32,32,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,64,0.005000000198682149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,32,0.004936000125275718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,512,0.005296000176005893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,128,0.004969777746333016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,768,32,768,0.0035386664999855887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,16384,0.77255556318495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,6144,0.013062222136391534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,16384,0.3092950979868571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,16384,0.7461742295159234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,51200,2.407348420884874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,1024,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,10240,0.43887911902533633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,12288,0.23250667254130045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,51200,0.9629191292656792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,12288,0.5522115495469835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,10240,0.2057804399066501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,8192,0.3533715671963162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,7168,0.30760622024536133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,8192,0.16151732868618435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,8192,0.3611884382035997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,7168,0.14298933082156712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,7168,0.3224995666080051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,6144,0.12480089399549697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,6144,0.27683644824557835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,5120,0.2325537734561496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,5120,0.10586933294932048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,10240,0.45592355728149414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,4096,0.08749066458808051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,4096,0.18672088781992593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,6144,0.2640026675330268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,3072,0.14054311646355525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,5120,0.2387777699364556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,3584,0.16159466902414957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,2560,0.1123537752363417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,51200,2.594926198323568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,3072,0.14079556200239393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,2560,0.058782219886779785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,2048,0.09996710883246528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,3584,0.15402044190300837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,1536,0.07370577918158637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,2048,0.05113244387838575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,2048,0.09969155655966865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,1536,0.04036888811323378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,1024,0.049794667296939425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,1536,0.07867377996444702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,4096,0.21539199352264404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,768,0.04055733482042948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,1024,0.031074666314654883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,1024,0.05929778019587199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,768,0.02846222122510274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,2560,0.1207502285639445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,256,0.023330666952663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,512,0.04214222232500712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,3584,0.07922310961617364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,128,0.020495999190542433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,256,0.02388444377316369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,3072,0.06887644529342651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,128,0.023754666248957317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,64,0.01789777808719211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,512,0.031011554929945204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,12288,0.5771510865953233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,768,0.048919111490249634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,64,0.03460266523891025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,512,0.026376888155937195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,32,0.03461511267556085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,256,0.03716533382733663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,128,0.03459466828240289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,32,0.01805688937505086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,16384,0.6068062252468532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,16384,0.5801537831624349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,65536,1.0578062269422743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,16384,0.2493351035647922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,65536,2.45230844285753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,51200,2.0128337012396917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,65536,2.46196534898546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,10240,0.4033031198713515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,12288,0.19122666782803008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,51200,1.8946747250027125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,10240,0.15679732958475748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,8192,0.2963502142164442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,10240,0.35321511162651914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,51200,0.8034248881869845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,7168,0.2753431002298991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,12288,0.4113910992940267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,8192,0.2897128793928358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,6144,0.20669510629442003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,7168,0.11333689424726699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,6144,0.10334666570027669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,5120,0.17093155119154188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,6144,0.21667644712660047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,5120,0.08619733651479085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,5120,0.18173066775004068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,3584,0.13803911209106445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,4096,0.07118311193254259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,4096,0.14646044042375353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,3072,0.1089022225803799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,3584,0.06326311164432101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,3584,0.12792000505659315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,7168,0.25006045235527885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,3072,0.05566666523615519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,4096,0.1401653289794922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,2560,0.04859822326236301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,2048,0.09643733501434326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,2560,0.09575911362965901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,1536,0.07439644469155206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,2048,0.04083377785152859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,2048,0.07925688558154635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,1024,0.05120444297790527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,1536,0.06298310889138116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,1024,0.04799555407630073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,768,0.035220444202423096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,12288,0.4413253201378717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,512,0.027471999327341717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,768,0.038972444004482694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,512,0.02129777769247691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,256,0.020436444216304354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,512,0.0327413347032335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,256,0.019440889358520508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,1024,0.025747555825445387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,128,0.016509332590632968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,128,0.019375110665957134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,768,0.02310755517747667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,64,0.014816888504558139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,32,0.014512888259357877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,64,0.027946667538748846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,8192,0.12731556097666422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,32,0.027296889159414504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,2560,0.11148444811503093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,256,0.029250668154822454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,65536,0.6693386501736112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,3072,0.11111999882592095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,51200,0.5314568943447536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,65536,0.7743848694695367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,1536,0.03321155574586656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,16384,0.0883751114209493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,12288,0.15853599707285562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,16384,0.1895057757695516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,10240,0.11978044774797227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,128,0.02774311105410258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,12288,0.06653955247667101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,65536,0.34948354297214085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,8192,0.10053510798348321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,10240,0.12023554907904731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,7168,0.08623822530110677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,8192,0.046599109967549644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,8192,0.09743289152781169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,51200,0.2664026684231228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,6144,0.0786071088578966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,7168,0.08579555484983657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,6144,0.03666844301753574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,12288,0.14203643798828125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,6144,0.07536800040139092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,4096,0.05479022198253208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,5120,0.031874666611353554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,16384,0.2082444429397583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,5120,0.06397600306404962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,3584,0.04725244310167101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,51200,0.6268711090087891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,4096,0.052546666728125684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,3584,0.024105777343114216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,3584,0.04657155606481764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,7168,0.041984889242384166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,2560,0.03547022077772353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,3072,0.021916444102923077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,3072,0.041089776489469744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,5120,0.06847644514507718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,2560,0.01975288821591271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,1536,0.02350844442844391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,2560,0.035805334647496544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,2048,0.01755822201569875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,1024,0.017318222257826064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,1536,0.015210666590266757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,1536,0.024442666106753882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,768,0.01464355488618215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,3072,0.045632888873418175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,768,0.011586666935020022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,512,0.01239377756913503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,2048,0.029343111647499934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,512,0.011035555766688453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,512,0.013853333062595792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,10240,0.05700711409250895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,4096,0.02698755595419142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,1024,0.012812444733248817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,1024,0.01943288909064399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,128,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,256,0.010328888893127441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,64,0.0068195557428730865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,128,0.01034311122364468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,128,0.012151999606026543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,256,0.008934222161769867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,2048,0.029663999875386555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,32,0.012090666426552666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,256,0.012431999875439538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,65536,0.49238400989108616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,65536,0.2478382322523329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,32,0.006728000111050076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,51200,0.4070719877878825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,16384,0.15121511618296304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,51200,0.19668889045715332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,64,0.012118221984969245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,16384,0.0637520021862454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,16384,0.15971110926734075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,10240,0.10023644235399033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,12288,0.04936000042491489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,12288,0.10940888855192397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,8192,0.08213333288828532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,10240,0.042966223425335355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,7168,0.07028533352745904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,65536,0.6119333373175727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,8192,0.03466577662361993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,8192,0.07550933625963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,6144,0.06156177653206719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,7168,0.03421688742107815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,51200,0.4963066842820909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,6144,0.027826666831970215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,12288,0.11115377479129368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,6144,0.05882133377922905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,4096,0.042556444803873696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,5120,0.024358222881952923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,5120,0.05017422305213081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,4096,0.021143999364640977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,4096,0.041321777635150485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,3072,0.03070577647950914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,3584,0.03642311029964023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,3072,0.01772177716096242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,3072,0.03253511256641812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,7168,0.06703022453520033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,2560,0.016366221838527255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,2048,0.022100443641344707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,2560,0.02832177612516615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,768,0.016002666619088914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,2048,0.014522666732470194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,1536,0.01717688971095615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,2048,0.02388266722361247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,1024,0.01310400002532535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,1536,0.012553777959611682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,1536,0.020177778270509508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,768,0.010875555376211802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,1024,0.016024889217482675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,768,0.010089777410030365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,768,0.013165333204799227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,512,0.00850755555762185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,3584,0.02089599933889177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,256,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,1024,0.010712888505723743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,256,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,128,0.005592888842026393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,3584,0.03716355562210083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,128,0.00903200027015474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,10240,0.09211199813418919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,64,0.0052355556852287715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,512,0.01145155562294854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,32,0.005249777601824866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,64,0.010374222364690568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,5120,0.0488737788465288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,65536,0.5476391050550673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,65536,0.22604089313083223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,65536,0.5028924412197536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,512,0.009371555513805812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,51200,0.4418711132473416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,16384,0.14399822552998862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,51200,0.1821662187576294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,2560,0.027072888281610277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,16384,0.12374488512674968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,12288,0.11029689179526435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,10240,0.09699200259314643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,12288,0.09325866566763984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,128,0.01038933296998342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,10240,0.04052088989151849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,8192,0.07747289207246569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,8192,0.03287377622392442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,256,0.010621333287821876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,16384,0.061943113803863525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,7168,0.02966044346491496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,7168,0.057722667853037514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,10240,0.07889688677257962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,6144,0.05069244570202298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,5120,0.04807555675506592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,6144,0.027415111660957336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,32,0.010437332921557956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,51200,0.40600177976820206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,5120,0.023223999473783705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,4096,0.04001244571473863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,5120,0.04314666655328539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,3584,0.03460177779197693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,4096,0.02020355562369029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,6144,0.057856890890333385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,4096,0.03523911039034525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,3584,0.018728888697094388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,3584,0.03165422214402093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,2560,0.025390222668647766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,3072,0.017143999536832173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,2560,0.015586665934986539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,7168,0.06701599889331393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,2048,0.021353777911927965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,12288,0.04801599846945869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,1536,0.017050666941536795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,2048,0.021128000484572515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,3072,0.02981511089536879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,1024,0.012923555241690742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,1536,0.012039999994966718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,1536,0.017937777770890128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,1024,0.010355555348926121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,768,0.011148444480366178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,1024,0.014304000470373364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,512,0.00832622249921163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,8192,0.0646302236451043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,2560,0.02482844392458598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,256,0.006949333681000604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,512,0.009362666971153682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,2048,0.01389244364367591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,512,0.01038933296998342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,128,0.00562755556570159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,256,0.008767111433876885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,64,0.005006222261322869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,256,0.009691555466916826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,32,0.005223999834722943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,128,0.009538667069541084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,64,0.009475555684831407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,32,0.009694221946928237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,65536,0.36220622062683105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,768,0.009686222506894005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,51200,0.2916195657518175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,65536,0.17882577578226724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,65536,0.4156044324239095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,16384,0.10461244318220352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,51200,0.14152799712287054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,51200,0.33473867840237087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,16384,0.05066577924622429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,128,0.00889244427283605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,10240,0.06725777520073785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,12288,0.03827910953097873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,8192,0.05561421977149116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,10240,0.032681776417626276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,10240,0.0658586687511868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,7168,0.04698400033844841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,8192,0.028647998968760174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,768,0.012048888537618848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,8192,0.05389688743485344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,6144,0.04273866613705953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,7168,0.024494222468800012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,5120,0.03535111082924737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,12288,0.07632978094948663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,7168,0.048058665461010404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,6144,0.042111112011803516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,16384,0.10132355822457208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,5120,0.03633600142267015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,3072,0.02840355700916714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,4096,0.030020445585250854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,3584,0.016179554992251925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,3584,0.02696711156103346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,6144,0.022105778257052105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,3072,0.024192000428835552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,5120,0.019827556278970506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,12288,0.07763555314805773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,2560,0.02126844392882453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,2560,0.013739555246300168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,3072,0.02199288871553209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,2048,0.012297777665985955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,1536,0.013705777625242868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,4096,0.02811644474665324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,3072,0.015450666348139444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,1024,0.01071999967098236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,1536,0.010773333410422007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,4096,0.017321777012613084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,768,0.009394666386975182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,1024,0.009356444080670675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,2048,0.0162124451663759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,512,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,768,0.010495111346244812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,2048,0.018224888377719455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,512,0.008350222474998897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,256,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,2560,0.019235556324323017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,128,0.005247999810510212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,256,0.00871288859181934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,1536,0.015464888678656684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,64,0.00489688871635331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,128,0.008012444608741337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,1024,0.012076444096035428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,32,0.004899555610285865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,768,0.008784888519181145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,64,0.008743111458089616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,32,0.00868266655339135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,512,0.009381333159075843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,256,0.008030222521887885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,51200,0.2736097706688775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,3584,0.025567111041810777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,65536,0.16940089066823324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,16384,0.10142755508422852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,128,0.008671110702885522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,51200,0.14265955819023982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,16384,0.048287110196219556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,16384,0.09034488598505656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,10240,0.058760888046688504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,12288,0.03685422076119317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,12288,0.06963200039333768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,8192,0.05067200130886502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,10240,0.03205777870284186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,10240,0.05878133243984646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,65536,0.36352001296149355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,7168,0.04330044322543674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,8192,0.027606222364637587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,8192,0.04858666658401489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,7168,0.023856000767813787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,12288,0.07627289162741767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,7168,0.04285600119166904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,51200,0.2841164535946316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,6144,0.02180888917711046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,6144,0.03786311215824551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,5120,0.019582221905390423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,3584,0.023003555006451074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,5120,0.0324906673696306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,4096,0.017086222767829895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,4096,0.027249778310457867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,3584,0.015825778245925903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,3584,0.02460266649723053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,2560,0.017263111140992906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,3072,0.022095999783939783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,5120,0.03317155440648397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,2048,0.01481066644191742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,65536,0.334112008412679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,3072,0.02052888936466641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,1536,0.012496000362767113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,3072,0.014514666464593677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,1536,0.013931555880440606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,1536,0.010439111126793755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,4096,0.026291555828518335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,1024,0.00907466643386417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,768,0.008348444269763099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,2048,0.016794666647911072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,768,0.009972444838947719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,6144,0.0391591125064426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,512,0.006341333190600078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,1024,0.010030222435792288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,512,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,256,0.005568000177542369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,2560,0.019352000620630052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,128,0.004594666676388847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,256,0.007745777567227681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,1024,0.011083555718262991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,256,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,128,0.00793511089351442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,32,0.004570666700601578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,64,0.008032889001899296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,768,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,32,0.008048888709810045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,2560,0.01313777764638265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,512,0.008900444540712569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,51200,0.22681244214375815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,65536,0.3173911041683621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,64,0.00425511101881663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,128,0.00813244448767768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,51200,0.2558728853861491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,12288,0.0599857767422994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,16384,0.07965244187249078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,65536,0.23822487725151908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,12288,0.061627553568945996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,65536,0.13475110795762804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,16384,0.0812328921424018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,10240,0.05230755607287089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,8192,0.04273244407441881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,2048,0.011754666765530905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,7168,0.02959200077586704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,8192,0.02074399921629164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,16384,0.03642044464747111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,8192,0.04272177815437317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,6144,0.029878222280078467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,7168,0.019137778215938143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,12288,0.02857422166400486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,10240,0.04311199982961019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,6144,0.03370133373472426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,4096,0.020434666011068556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,5120,0.02920088834232754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,10240,0.025985777378082275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,3584,0.018405333161354065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,4096,0.015023110641373528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,4096,0.024549333585633173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,3072,0.015876443849669564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,3584,0.013154666456911298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,3584,0.022102221846580505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,5120,0.026378666361172993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,3072,0.012442666623327466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,6144,0.01752622259987725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,5120,0.015953777564896476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,2560,0.010957332948843638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,2048,0.011701333026091257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,51200,0.10669777790705363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,1536,0.010431999961535135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,2048,0.015164444843928019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,1536,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,1536,0.012774222426944308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,3072,0.020071999894248117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,2560,0.01382844481203291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,768,0.006796444455782573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,1024,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,1024,0.01038577738735411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,2560,0.017490666773584154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,768,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,768,0.009224889179070791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,2048,0.009744889206356471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,512,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,512,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,128,0.004232888834344016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,256,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,256,0.007688889073001013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,1024,0.007982222570313348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,128,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,128,0.007666666474607255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,64,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,32,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,65536,0.28789954715304905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,51200,0.22571288214789498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,65536,0.12518844339582655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,65536,0.27099021275838214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,64,0.003919110943873723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,32,0.0042151109211974675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,51200,0.21783644623226592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,16384,0.06898488601048787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,256,0.004606222112973531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,12288,0.05299733413590325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,12288,0.027552889453040227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,10240,0.045271999306148954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,7168,0.03789510991838243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,8192,0.03783911135461595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,51200,0.09840799702538384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,8192,0.02295822236273024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,8192,0.03682400120629205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,7168,0.03490044342146979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,7168,0.019202666150199044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,6144,0.030227555168999568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,16384,0.07501777675416735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,10240,0.024467556013001338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,5120,0.024533333049880132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,16384,0.03690133492151896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,5120,0.015784000356992085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,5120,0.025742222865422566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,4096,0.020563556088341605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,7168,0.033019555939568415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,4096,0.01402933398882548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,3584,0.018170666363504197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,4096,0.021869333253966436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,3072,0.016149333781666227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,3584,0.012773333324326409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,10240,0.04491466614935133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,3584,0.019907555646366544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,2560,0.014350222216712104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,3072,0.011690666278203329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,512,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,3072,0.017955554856194388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,2560,0.01071466671095954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,1536,0.010346666806273991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,2048,0.009705777797434065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,2048,0.013385777672131857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,1024,0.00796088907453749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,1536,0.01110222190618515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,1024,0.0081404447555542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,6144,0.029566221767001685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,768,0.006945778098371293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,6144,0.017525333497259352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,512,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,768,0.007977777885066139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,768,0.008356444537639618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,2560,0.015769778026474845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,512,0.007449777589903937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,512,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,128,0.00480355570713679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,256,0.007096000015735626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,256,0.0070364442136552595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,128,0.007270221908887227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,128,0.007326222128338284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,32,0.00425155543618732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,64,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,2048,0.012633778154850006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,32,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,256,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,65536,0.24191376898023817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,51200,0.1686800056033664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,1536,0.009485333330101436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,65536,0.22777067290412056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,16384,0.05417155557208591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,64,0.00424888895617591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,51200,0.17600622442033556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,51200,0.09359911415312026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,1024,0.009354666703277165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,16384,0.034549332327312894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,16384,0.05788889196183947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,12288,0.04457599918047587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,10240,0.023949333363109167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,8192,0.033041778537962176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,65536,0.11727110544840495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,8192,0.020000888241661918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,8192,0.03166755702760484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,7168,0.02760355505678389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,12288,0.02698755595419142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,7168,0.018581334087583754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,6144,0.024135111106766596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,7168,0.02849777870708042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,5120,0.02082577844460805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,6144,0.01699733402993944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,10240,0.0379342238108317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,6144,0.025407110651334126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,4096,0.016391999191708036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,5120,0.01514311134815216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,5120,0.0222080002228419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,4096,0.013302221894264221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,4096,0.018924444913864136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,12288,0.04660711023542616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,3584,0.012492444780137805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,3584,0.0174026671383116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,2560,0.012433778080675336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,3072,0.01144177797767851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,3072,0.015862221519152325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,2048,0.010408889088365767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,2560,0.010079999764760336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,1536,0.00905866672595342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,2048,0.01145066652033064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,3584,0.015615999698638916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,1024,0.006949333681000604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,1536,0.010059555371602377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,3072,0.01349955548842748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,768,0.006270222365856171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,1024,0.008410666551854875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,768,0.007636444436179266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,512,0.0053031109273433685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,2560,0.013184000220563678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,512,0.007329777710967594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,512,0.0070462218589252895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,1536,0.008681777450773451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,256,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,256,0.006635555376609166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,128,0.006953777538405524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,768,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,64,0.003571555432346132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,128,0.006666666517655055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,2048,0.009389333426952362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,64,0.006670222100284364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,256,0.00423911131090588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,1024,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,65536,0.11593511369493271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,65536,0.20937955379486084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,51200,0.14640355110168457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,16384,0.04960799879497952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,51200,0.09239377578099568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,32,0.003617777592606015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,51200,0.16775822639465332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,16384,0.032904889848497175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,65536,0.17884800169203016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,10240,0.03134844369358487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,12288,0.026416889495319788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,12288,0.04291911257637871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,128,0.004233777936961916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,10240,0.02360711163944668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,10240,0.036899556716283165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,7168,0.02416444487041897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,8192,0.019834667444229126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,8192,0.030846221579445735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,12288,0.042005333635542125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,7168,0.018184888694021437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,16384,0.056051552295684814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,5120,0.018552000323931377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,6144,0.016604445046848722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,6144,0.024863999750879075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,4096,0.015990222493807476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,10240,0.03994489047262404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,8192,0.026927111877335444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,3584,0.01443288889196184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,6144,0.021354667014545862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,3584,0.016883555385801528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,3584,0.012094222009181976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,32,0.006717333363162146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,3072,0.011111111276679568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,2560,0.01108444482088089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,5120,0.021738666627142165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,2560,0.010028444230556488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,2048,0.010370666782061258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,7168,0.027818666564093694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,2048,0.011415999796655444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,1536,0.008661333057615492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,2048,0.009332444104883406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,5120,0.01475911173555586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,3072,0.015326221783955893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,1024,0.006606222026877933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,1536,0.008620444271299574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,4096,0.012786666552225748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,2560,0.013165333204799227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,1024,0.00832444429397583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,768,0.007748444047239091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,3072,0.012843555874294706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,512,0.0070062221752272705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,1536,0.009927999642160203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,256,0.004222222086456087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,512,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,768,0.005718222094906702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,128,0.003899555653333664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,256,0.006981333096822103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,64,0.003540444291300244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,128,0.006959999601046245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,512,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,128,0.0063075555695427795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,64,0.0063271112740039825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,65536,0.19234489070044625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,32,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,51200,0.1514337725109524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,4096,0.018609777092933655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,1024,0.00796622203456031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,16384,0.04836533466974894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,51200,0.060722668965657554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,51200,0.14005333847469753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,768,0.007679111427730984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,16384,0.04639110962549845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,12288,0.018580444984965853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,12288,0.03580177823702494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,10240,0.03101955519782172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,65536,0.07688977983262804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,10240,0.016899555921554565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,8192,0.025049777494536504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,10240,0.03085422184732225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,7168,0.02186044388347202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,256,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,8192,0.026145777768558923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,12288,0.03754222061898973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,6144,0.019370666808552213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,7168,0.023543111152119104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,5120,0.016895999511082966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,6144,0.021191999316215515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,5120,0.011361777782440186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,65536,0.179495996899075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,4096,0.014483556151390076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,8192,0.017208000024159748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,3584,0.013110222087966071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,16384,0.0225564440091451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,7168,0.013468444347381592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,3072,0.011719110939237805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,3584,0.010412444670995077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,3584,0.014536889062987434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,2560,0.010408889088365767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,3072,0.008995555341243744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,3072,0.013160000244776407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,2048,0.00904799997806549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,2560,0.008663111262851292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,2560,0.011404444773991903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,2048,0.008227555288208855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,4096,0.010062221851613786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,2048,0.010223111344708337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,4096,0.01590044465329912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,1024,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,1536,0.007650666766696506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,768,0.005239999956554837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,1024,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,6144,0.012093333734406365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,512,0.004837333328194088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,768,0.007003555695215861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,5120,0.01870577699608273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,256,0.004239999999602635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,1536,0.007754666937722101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,32,0.003546666767862108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,256,0.006457777900828256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,128,0.0038817777401871155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,1536,0.008711111214425828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,64,0.003528888854715559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,768,0.006851555572615729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,32,0.00360977773865064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,64,0.006102222121424145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,512,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,32,0.0063075555695427795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,256,0.006311999840868845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,65536,0.06963022549947102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,65536,0.16818488968743217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,51200,0.13616622818840876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,128,0.0063902222447925145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,16384,0.04548799991607666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,128,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,51200,0.056523554854922824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,51200,0.13134845097859701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,12288,0.03659822212325202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,10240,0.03047200043996175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,12288,0.01792622274822659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,12288,0.034837332036760114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,65536,0.17496711677975127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,8192,0.024243555135197107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,10240,0.030596445004145306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,8192,0.013844444519943662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,8192,0.02555911077393426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,16384,0.021497777766651575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,7168,0.023192889160580103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,6144,0.019076445036464267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,7168,0.012844444149070315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,10240,0.01622577839427524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,6144,0.01145244472556644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,5120,0.016605334149466622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,6144,0.020844444632530212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,4096,0.014474666780895658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,5120,0.010497777826256223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,1024,0.007670222057236566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,5120,0.018227555685573153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,3584,0.013455111119482251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,4096,0.009687999884287516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,4096,0.015268445014953613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,3584,0.009382222261693742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,16384,0.045069333579805165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,512,0.006746666712893381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,3072,0.008815999660227034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,3072,0.012521777715947894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,2048,0.009347555538018545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,2560,0.008325333396593729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,2560,0.011186666786670685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,2048,0.008024888734022776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,7168,0.022608000371191237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,1536,0.0076657773719893555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,1536,0.008681777450773451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,3072,0.011768889096048145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,768,0.00553688903649648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,1024,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,1024,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,512,0.004559111264016894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,768,0.006797333144479328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,768,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,512,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,512,0.006635555376609166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,1536,0.007707555260923173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,128,0.003552000141806073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,256,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,256,0.006149333384301927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,64,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,128,0.006308444258239534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,32,0.003572444534964032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,3584,0.013809777796268463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,2560,0.01053244454993142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,64,0.005976000179847081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,65536,0.16827732986874053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,32,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,51200,0.1314986679289076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,65536,0.13184977902306452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,51200,0.05474311113357544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,51200,0.10389955838521321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,1024,0.006002666635645761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,12288,0.03405066662364536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,16384,0.021003555920388963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,128,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,16384,0.035206221871905856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,10240,0.028547555208206177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,12288,0.028010666370391846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,10240,0.01571466690964169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,10240,0.024562666813532513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,8192,0.013264888690577613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,16384,0.04103111227353414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,7168,0.02032088902261522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,7168,0.012479999827014076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,6144,0.017843556072976854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,12288,0.017516444126764934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,8192,0.022820444570647344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,6144,0.017044444878896076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,2048,0.00998399986161126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,5120,0.014671110444598727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,4096,0.012903110848532783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,7168,0.018534221582942538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,4096,0.012470222181744046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,3584,0.012127999630239276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,6144,0.01143022212717268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,5120,0.015882665912310284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,3072,0.01071466671095954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,3584,0.009189333352777693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,3584,0.011791110866599612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,8192,0.020635555187861126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,2560,0.009331555830107795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,3072,0.010724444356229572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,5120,0.010375111467308467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,2048,0.008676444490750631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,2560,0.009644444617960189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,1536,0.0069520001610120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,4096,0.009704888694816166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,256,0.003937777959638172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,1536,0.007682667010360294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,1024,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,3072,0.008683555656009251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,768,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,1024,0.0069573331210348345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,2560,0.008360000120268928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,768,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,512,0.004234666625658671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,2048,0.008633777499198914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,512,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,512,0.005964444329341252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,256,0.0038986665507157645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,256,0.005623999983072281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,256,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,2048,0.00777688870827357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,128,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,128,0.006274666637182236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,65536,0.06785600052939521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,32,0.0031804444475306403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,64,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,32,0.005636444522274866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,65536,0.10548710823059082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,128,0.0035084444615576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,51200,0.08340266678068374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,768,0.0062977779242727495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,16384,0.03223289052645365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,51200,0.0553342236412896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,64,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,1536,0.007325333025720384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,16384,0.021311110920376245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,16384,0.034781333472993635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,10240,0.022543110781245764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,12288,0.027503111296229895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,10240,0.01521511044767168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,65536,0.06677777899636163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,65536,0.12705866495768228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,8192,0.012755555411179861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,8192,0.02032800018787384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,7168,0.01683288812637329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,12288,0.017308443784713745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,6144,0.015142222245534262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,10240,0.024088000257809956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,8192,0.018737778067588806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,6144,0.016302223006884258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,5120,0.01332355539004008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,5120,0.010300444232092964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,12288,0.026179555389616225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,4096,0.011415111521879831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,4096,0.009362666971153682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,7168,0.011857777833938599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,3584,0.01036711119943195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,7168,0.018231110440360177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,3072,0.009356444080670675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,3584,0.009006222089131674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,3584,0.011622221933470832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,2560,0.008031110796663497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,6144,0.011074666347768573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,3072,0.01072266697883606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,2048,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,2560,0.008325333396593729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,51200,0.09865422381295098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,1536,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,2048,0.007713778151406183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,2048,0.008671999805503422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,1024,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,1536,0.007336888876226213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,1536,0.007675555845101674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,768,0.004581333448489507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,1024,0.0069351109365622205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,1024,0.006699555450015598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,3072,0.008432000047630733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,512,0.004253333227501975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,768,0.006317333214812809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,256,0.003568888952334722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,2560,0.009573332965373993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,256,0.006262222097979651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,256,0.005640888793600931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,4096,0.012416889270146688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,128,0.006309333360857434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,64,0.003206222214632564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,768,0.00665155549844106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,32,0.0033475555893447665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,64,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,5120,0.014480888843536377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,65536,0.07135111093521118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,32,0.005595555735958948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,128,0.003536888708670934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,51200,0.056008888615502246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,65536,0.08176000250710382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,16384,0.02164177762137519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,51200,0.0521048903465271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,128,0.0053386667536364655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,51200,0.0649777783287896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,12288,0.018349332941903006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,16384,0.01995022263791826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,10240,0.019253333409627277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,12288,0.01650044487582313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,12288,0.019478221734364826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,8192,0.01514311134815216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,1024,0.0070159998204973005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,65536,0.0653066635131836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,7168,0.014463111758232117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,8192,0.012643555800120035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,7168,0.01179377817445331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,6144,0.012306667036480375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,16384,0.024146666129430134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,512,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,512,0.005965333431959152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,5120,0.011705777711338468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,6144,0.010773333410422007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,10240,0.01718488832314809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,4096,0.009649777577983009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,5120,0.010064889159467485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,8192,0.014478221535682678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,3584,0.00905600024594201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,4096,0.0092613332801395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,4096,0.009862222605281407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,7168,0.01349866638580958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,3584,0.008864888714419471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,3584,0.009344889058007134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,2560,0.0069511110583941145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,6144,0.012155555188655853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,10240,0.0148062225845125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,2048,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,2560,0.007927999728255803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,5120,0.011085333095656501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,1536,0.005586666779385672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,3072,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,1536,0.00665155549844106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,3072,0.008462222086058723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,3072,0.00867377801073922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,2560,0.008024888734022776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,1024,0.005962666538026597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,2048,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,512,0.003882666842805015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,768,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,768,0.005843555761708154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,256,0.0032666667054096856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,512,0.006274666637182236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,512,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,1536,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,256,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,256,0.005250666704442766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,64,0.0029084444459941653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,128,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,32,0.0031973332580592898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,64,0.005232000102599462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,2048,0.007244444555706448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,768,65536,0.0551111102104187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,65536,0.06469066937764485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,65536,0.07957688967386882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,768,51200,0.044300443596310086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,51200,0.051170667012532554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,128,0.005047111047638787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,32,0.0052800000541739995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,16384,0.023527999718983967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,768,12288,0.015863999724388123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,16384,0.019136889113320243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,1024,0.005205333232879639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,768,10240,0.014499555031458536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,12288,0.018935999936527677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,1024,0.006643555644485686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,768,0.004249777644872665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,10240,0.014493332968817817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,10240,0.016522667474216886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,768,7168,0.01481066644191742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,8192,0.014192889134089151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,7168,0.01180622229973475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,768,6144,0.011249778171380361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,12288,0.016192000773217943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,6144,0.010778666370444827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,51200,0.06285511122809516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,768,5120,0.009677333136399588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,8192,0.012433778080675336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,5120,0.010084444450007545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,768,4096,0.008358221915033128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,5120,0.010785777535703448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,768,3584,0.007675555845101674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,4096,0.009355555805895064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,7168,0.013119999733236102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,4096,0.009687999884287516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,3584,0.00869422240389718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,3584,0.0090506664580769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,768,2560,0.006640000061856375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,3072,0.008349333372380998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,768,8192,0.015835555063353646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,768,2048,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,2560,0.008021333151393467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,2560,0.007697777615653143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,768,1536,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,2048,0.0076657773719893555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,2048,0.006991110742092133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,768,1024,0.0042151109211974675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,1536,0.007297777467303806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,768,3072,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,768,768,0.004212444441186057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,1024,0.006743999818960826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,1024,0.006047111004590988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,768,512,0.003548444559176763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,768,16384,0.018216888109842937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,512,0.006310222049554189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,512,0.005442666510740916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,256,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,256,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,3072,0.008348444269763099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,768,128,0.0028977776981062363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,768,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,6144,0.011888888974984487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,768,64,0.0032248888164758682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,128,0.006311111152172089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,128,0.005047111047638787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,768,32,0.0031893334041039148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,768,256,0.0033235556135574975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,512,65536,0.039988444911109075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,64,0.005129777722888523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,32,0.004948444250557157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,65536,0.06373333268695407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,65536,0.07856888903511895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,512,16384,0.01480888823668162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,51200,0.049775110350714795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,51200,0.060715556144714355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,512,12288,0.014817777607176038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,16384,0.01898755629857381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,16384,0.023012444376945496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,512,10240,0.013150222599506378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,12288,0.015856888559129503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,12288,0.018396443790859647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,512,8192,0.0138586668504609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,10240,0.014478221535682678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,512,51200,0.03218844532966614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,128,0.0032382222513357797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,8192,0.01242844429281023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,8192,0.013821333646774292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,512,6144,0.010394666757848527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,7168,0.012757333616415659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,768,0.005655111124118169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,512,5120,0.009391110804345872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,6144,0.01071733319097095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,5120,0.010048888623714447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,5120,0.010457778142558204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,10240,0.016001777516471017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,1536,0.006682666639486949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,512,3584,0.007336888876226213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,4096,0.00941422250535753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,512,3072,0.006606222026877933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,3584,0.00870222184393141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,3584,0.009020444419648912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,512,2560,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,3072,0.008396444221337637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,3072,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,512,2048,0.005292444593376583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,2560,0.008012444608741337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,512,4096,0.007975111405054728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,512,1536,0.004777777940034866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,2048,0.007698666718271043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,2048,0.006960888703664144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,512,1024,0.00423555572827657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,1536,0.007292444507280986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,4096,0.009180444810125563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,1536,0.006597333484225803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,512,768,0.003875555677546395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,1024,0.006953777538405524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,512,512,0.003565333369705412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,768,0.00665155549844106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,6144,0.011751111182901593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,768,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,512,256,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,512,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,512,128,0.0030497778207063675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,512,0.0052933332820733385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,256,0.006252444452709622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,512,64,0.002879111096262932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,512,7168,0.012095111111799875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,128,0.004935111022657818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,7168,0.01146044416560067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,256,65536,0.02572088936964671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,1024,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,65536,0.07680088943905301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,65536,0.06067733632193672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,256,51200,0.022218666142887537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,2560,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,256,0.005286222116814719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,128,0.006269333263238271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,51200,0.04849333233303494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,51200,0.05882133377922905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,16384,0.018784888916545443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,64,0.004934222333961063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,256,10240,0.009848000274764167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,12288,0.01588800052801768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,12288,0.01789866718981001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,256,8192,0.010382222632567087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,10240,0.014515555567211576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,10240,0.015871110889646742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,256,7168,0.009726222190592024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,8192,0.0136835558546914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,256,16384,0.012088889049159156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,256,6144,0.009346666435400644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,7168,0.011416888899273343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,512,32,0.0030097777230872046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,256,5120,0.008662222160233391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,16384,0.022100443641344707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,6144,0.010737777584128909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,6144,0.011495999991893768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,5120,0.009701333112186855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,32,0.004976888911591636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,256,3584,0.007288888924651676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,4096,0.00902844468752543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,4096,0.009359999663299983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,8192,0.01241422196229299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,256,3072,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,3584,0.008725333544943068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,256,12288,0.010833777487277985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,3072,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,256,2560,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,256,4096,0.007967111137178209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,2560,0.008001777860853408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,256,2048,0.0052391112678580815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,2560,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,5120,0.010454222559928894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,256,1536,0.004903111192915174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,2048,0.007638221813572778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,2048,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,1536,0.007228444847795699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,3584,0.008662222160233391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,256,768,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,1024,0.006653333289755716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,1024,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,7168,0.012513777448071374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,768,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,768,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,256,256,0.0031964443624019623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,512,0.0052782222628593445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,256,0.005993777679072485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,256,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,256,128,0.002993777808215883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,1536,0.006362666686375936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,128,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,128,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,256,32,0.002579555536309878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,128,65536,0.018965333700180054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,64,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,512,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,128,51200,0.01725955473052131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,65536,0.07617421944936116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,256,1024,0.0042275554604000514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,51200,0.04657155606481764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,128,16384,0.012098666694429187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,256,64,0.0028968888024489083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,16384,0.018575110369258456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,256,512,0.003540444291300244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,3072,0.008319999608728621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,128,10240,0.009377777576446533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,32,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,65536,0.058357331487867564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,10240,0.014386667145623101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,10240,0.01587199999226464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,51200,0.05883555279837715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,8192,0.012410666379663678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,8192,0.01348088930050532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,128,12288,0.010074666804737514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,16384,0.02201955517133077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,7168,0.012476444244384766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,12288,0.015576000014940897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,12288,0.017872888180944655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,128,5120,0.00831200016869439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,6144,0.011412444214026133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,128,4096,0.007661333514584436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,5120,0.009938666390048133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,5120,0.010416888528399997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,128,3584,0.007103111180994246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,4096,0.009356444080670675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,3584,0.00867733359336853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,3584,0.008780444661776224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,128,3072,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,3072,0.008402666283978356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,128,2560,0.0059164443777667145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,128,8192,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,2560,0.007979555262459649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,4096,0.009009777671760982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,128,2048,0.005290666802061929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,128,7168,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,128,1536,0.004561777744028303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,2048,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,7168,0.011420444481902652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,2048,0.006990222467316522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,128,1024,0.0041244443919923575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,1536,0.007089777953094906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,128,768,0.0038728887836138406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,1024,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,1024,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,128,512,0.0032151111712058387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,768,0.005623999983072281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,2560,0.007299555672539606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,512,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,512,0.005295111073387994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,128,6144,0.009026666482289633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,256,0.005992888990375731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,256,0.0053031109273433685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,1536,0.006311111152172089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,6144,0.010709332923094431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,128,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,3072,0.008249777886602614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,128,0.004945777770545748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,768,0.006395555618736479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,64,65536,0.016194666425387066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,64,51200,0.0135057775510682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,64,0.004996444616052839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,32,0.004960888789759742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,128,256,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,64,16384,0.009669333696365356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,65536,0.07489333550135295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,64,12288,0.01240799989965227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,64,10240,0.01109777804878023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,128,128,0.0028844444702068963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,16384,0.021957332889238994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,64,8192,0.010103111465771994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,64,7168,0.009302222066455418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,128,64,0.0028479999552170434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,128,32,0.0028728888266616394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,64,6144,0.008643555144468943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,8192,0.013461333182122974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,7168,0.012456888953844706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,64,4096,0.007638221813572778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,6144,0.011402666568756104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,64,3584,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,51200,0.05762400229771932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,4096,0.009352888498041365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,3584,0.00869599978129069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,64,2560,0.0058266665372583605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,10240,0.015623999966515435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,3072,0.008209777375062307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,64,2048,0.004898666507667965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,64,1536,0.004566222015354368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,64,1024,0.0038720000949170855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,2048,0.006691555596060223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,1536,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,5120,0.010372444159454769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,1024,0.0060159998635451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,64,3072,0.006305777778228124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,768,0.005587555468082428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,12288,0.017550221747822232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,64,128,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,256,0.005079111291302575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,64,64,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,128,0.004872888740566042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,64,32,0.0028728888266616394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,64,768,0.003546666767862108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,32,65536,0.036341332727008395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,32,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,64,256,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,512,0.005342222336265776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,32,16384,0.010448888772063784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,64,5120,0.008006222546100616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,2560,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,16384,0.02204088866710663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,32,10240,0.018566222654448617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,64,512,0.003253333270549774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,32,8192,0.016202666693263583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,32,51200,0.02847733431392246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,32,7168,0.013805333111021252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,65536,0.07459200090832181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,32,6144,0.0131217779384719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,8192,0.013556444810496436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,32,5120,0.011755555868148804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,32,12288,0.021354667014545862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,32,4096,0.009047110875447592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,6144,0.011426666544543372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,32,3584,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,5120,0.010386666489972008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,64,0.004980444494220945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,32,3072,0.007305777735180325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,4096,0.00925333301226298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,32,2560,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,32,2048,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,3072,0.008028444316652086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,2560,0.007522666619883643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,2048,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,32,1024,0.0038782221575578055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,7168,0.012435555458068848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,32,768,0.003536888708670934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,1024,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,12288,0.01757955551147461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,768,0.005442666510740916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,32,256,0.002990222225586573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,3584,0.008761777646011775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,32,128,0.0025253333151340485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,32,1536,0.004550222307443619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,32,64,0.002519999941190084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,51200,0.057568887869517006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,32,32,0.0025742221623659134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,1536,0.006270222365856171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,64,0.004949333353175057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,32,0.004970666848950916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,10240,0.015867556134859722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,512,0.005281777845488654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,256,0.004936000125275718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,128,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,16384,0.6104124387105306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,513,32,512,0.0032115555885765287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,16384,0.6722710927327474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,51200,2.1628345913357205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,12288,0.47066577275594074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,16384,0.23828532960679796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,51200,0.7130648824903699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,51200,2.2469003465440536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,12288,0.1781146658791436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,10240,0.4093146589067247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,12288,0.487427552541097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,8192,0.2999742296006945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,10240,0.50327820248074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,10240,0.151311993598938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,16384,0.590660466088189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,8192,0.12523999479081896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,8192,0.3250817722744412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,10240,0.3711208767361111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,12288,0.46127912733289933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,7168,0.2862311204274495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,8192,0.29990577697753906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,7168,0.10705066389507717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,6144,0.09479022026062012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,7168,0.26418312390645343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,7168,0.27748089366488987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,5120,0.20114133093092176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,6144,0.22919644249810112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,5120,0.20775555239783394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,6144,0.24794578552246094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,5120,0.1922231117884318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,4096,0.16737422678205702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,4096,0.07032355335023668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,3584,0.15260533491770426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,3584,0.14555822478400335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,3584,0.05973866913053724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,3072,0.13762666119469538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,3072,0.12728444735209146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,6144,0.24689245223999023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,4096,0.1569466723336114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,3584,0.1443679995006985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,2560,0.10982577668295966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,2560,0.04730133215586344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,2048,0.10011911392211914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,3072,0.12199910481770833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,2048,0.09099821911917792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,5120,0.0805564456515842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,2048,0.08591644631491767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,1536,0.07081244389216106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,2048,0.04047200083732605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,51200,2.0147760179307723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,1536,0.07184711429807875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,1024,0.053879999452167086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,1536,0.06958666774961683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,1024,0.026684444811609056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,1024,0.05448444353209602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,768,0.03800888856252035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,4096,0.17431733343336317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,1024,0.05399822195370992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,512,0.029667556285858154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,768,0.04499288731151157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,768,0.04558133416705661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,2560,0.10671733485327826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,256,0.020250666472646926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,512,0.038789331912994385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,2560,0.10354222191704644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,512,0.02162844439347585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,128,0.019519110520680744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,256,0.030496888690524634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,128,0.03154222170511881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,128,0.025715554753939312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,1536,0.03353777858946059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,64,0.016872889465755887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,32,0.01719288859102461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,3072,0.05299911234113905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,64,0.031465778748194374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,256,0.01958400011062622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,32,0.031063109636306763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,768,0.024111110303137038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,128,0.019480889042218525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,512,0.03842488924662272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,256,0.034244444635179304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,65536,2.207468456692166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,51200,1.686147583855523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,65536,2.1667662726508246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,51200,1.7457706663343642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,16384,0.47684446970621747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,51200,0.6040684382120768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,16384,0.5192889107598199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,16384,0.1845395565032959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,65536,0.756407101949056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,12288,0.40004534191555446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,12288,0.1409440040588379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,10240,0.3734773264990912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,16384,0.46285687552558047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,10240,0.1211191150877211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,12288,0.36951645215352374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,8192,0.26728977097405326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,12288,0.3505297766791449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,7168,0.21905956003401014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,8192,0.09615288840399848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,10240,0.3234213458167182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,7168,0.22123821576436362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,7168,0.08988889058430989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,6144,0.22004355324639213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,8192,0.25712354977925617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,10240,0.29230043623182506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,6144,0.1949368847741021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,5120,0.15879644287957087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,6144,0.07768355475531684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,8192,0.23584177758958605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,6144,0.18031022283766004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,5120,0.06421066655053033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,4096,0.14827822314368352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,7168,0.20897332827250162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,65536,2.007207022772895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,4096,0.13099822733137342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,3584,0.13389955626593694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,4096,0.12419466177622478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,51200,1.5714311599731445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,3584,0.04832977718777127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,3584,0.11531910631391738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,5120,0.16160444418589273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,3584,0.1148444414138794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,2560,0.0830577810605367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,3072,0.100746664736006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,3072,0.09557688898510402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,4096,0.0541928874121772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,2048,0.0753600001335144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,2560,0.03809244434038798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,2560,0.08258044719696045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,3072,0.1166799995634291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,2048,0.07057422399520874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,2048,0.07271377907858954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,5120,0.15233066346910265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,1536,0.027200889256265428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,1024,0.043656001488367714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,2560,0.0871644417444865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,1024,0.04418933391571045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,768,0.029451555675930444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,1024,0.04457155532307095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,2048,0.03308711118168301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,1536,0.06087288591596815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,512,0.02476177778508928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,768,0.018452444010310702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,768,0.03646755549642775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,1536,0.0586035582754347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,256,0.020319110817379422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,1536,0.057103998131222196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,1024,0.020622221959961783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,256,0.02676800058947669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,128,0.016876444220542908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,256,0.025106666816605464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,256,0.015562666787041558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,3072,0.04375999834802416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,64,0.015818667080667283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,128,0.01567911108334859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,512,0.01705333259370592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,32,0.01607733302646213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,512,0.030291554000642564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,512,0.0322746667597029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,64,0.025447110335032146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,32,0.025143999192449782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,65536,0.6087910864088271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,65536,0.26185067494710285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,65536,0.6835982004801432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,128,0.02202933364444309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,768,0.03798310955365499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,16384,0.16296622488233778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,51200,0.5706008805168999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,51200,0.20170044898986816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,16384,0.06816444132063124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,12288,0.13828266991509333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,51200,0.49208980136447483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,128,0.025401777691311304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,10240,0.11709777514139812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,12288,0.12912000550164118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,65536,0.6556151178148057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,12288,0.05244444476233589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,8192,0.09710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,10240,0.045132444964514844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,51200,0.5176062054104275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,10240,0.10625333256191677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,16384,0.17143021689520943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,8192,0.08737955490748088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,16384,0.1626960039138794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,7168,0.07780622111426459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,7168,0.03304444419013129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,7168,0.07747911082373725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,12288,0.12479733096228705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,6144,0.06812533405092028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,5120,0.051889777183532715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,8192,0.038916445440716214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,7168,0.07353422376844618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,8192,0.08767733309004043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,5120,0.024950222836600408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,4096,0.04846488767200046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,5120,0.057963556713528104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,5120,0.059451553556654185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,6144,0.06106488572226631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,4096,0.02146400014559428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,4096,0.0476471119456821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,4096,0.05037066671583387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,3072,0.03708622190687392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,3584,0.04233955674701267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,6144,0.06851644648445977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,3584,0.04524622360865275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,2560,0.031567109955681696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,3072,0.01794666714138455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,3072,0.04049688908788893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,2560,0.03240622083346049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,3584,0.045328888628217906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,2560,0.035389333963394165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,2048,0.027444443768925134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,10240,0.10775911145740086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,3584,0.019534221953815885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,2048,0.02723111046685113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,2048,0.030572444200515747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,3072,0.03763733307520548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,1536,0.02252799934811062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,1024,0.01555288831392924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,1536,0.012888000243239932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,2560,0.016529776983790927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,1024,0.017983999517228868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,1024,0.02122488949033949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,2048,0.014825777875052558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,768,0.01495733360449473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,768,0.018790221876568265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,768,0.010106667048401302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,512,0.010037333601050908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,512,0.01298577752378252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,768,0.01257511145538754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,256,0.007305777735180325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,512,0.009389333426952362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,256,0.009033777647548253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,128,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,256,0.012881778180599213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,1536,0.026258667310078938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,128,0.009024000002278222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,64,0.005695999910434087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,512,0.016007999579111736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,32,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,1536,0.020090666082170274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,64,0.011412444214026133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,256,0.011743110915025076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,32,0.011425777441925473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,65536,0.4649493429395888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,128,0.01145600030819575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,65536,0.5277066760592991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,65536,0.2134693331188626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,51200,0.3893244531419542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,1024,0.010987555815113915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,6144,0.028759999407662287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,16384,0.1223520040512085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,51200,0.16466222869025335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,51200,0.412172449959649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,16384,0.13355467054578993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,12288,0.10392710897657607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,16384,0.056017776330312095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,16384,0.12590932846069336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,12288,0.04211733407444424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,12288,0.09986755583021377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,128,0.012448888685968188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,65536,0.4879715707567003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,10240,0.03604355454444885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,10240,0.08488800128300984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,51200,0.39610666698879665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,10240,0.08489066362380981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,7168,0.07192711035410564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,8192,0.030080888006422255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,8192,0.06947644551595052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,7168,0.026662222213215295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,10240,0.08976711167229547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,12288,0.09714577595392863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,8192,0.07948178052902222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,5120,0.050613333781560264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,6144,0.023942222197850544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,6144,0.055052445994483105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,8192,0.06990844673580594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,5120,0.02038755516211192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,5120,0.04660977919896444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,6144,0.05563822057512072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,5120,0.047687109973695546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,7168,0.06413244538837008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,4096,0.03836266530884637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,4096,0.03993955585691664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,6144,0.05463822351561653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,3584,0.03417066733042399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,3584,0.016674665941132438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,3584,0.03661155700683594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,3072,0.0306346681382921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,4096,0.03755555550257365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,3072,0.03287555442916022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,7168,0.062257779969109424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,3584,0.03602044449912177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,4096,0.018394667241308425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,2560,0.029231998655531142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,3072,0.031409776873058744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,2048,0.022704000274340313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,1536,0.01626044511795044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,2048,0.025566221939192876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,3072,0.01531733406914605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,2560,0.027143999934196472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,1024,0.013187555803192986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,1536,0.010830221904648675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,1536,0.021735111872355144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,2048,0.02203111184967889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,768,0.011554666691356234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,1024,0.009377777576446533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,768,0.01237422227859497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,768,0.015680889288584392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,768,0.008889777792824639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,512,0.00905688852071762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,2560,0.014304000470373364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,1024,0.015136889285511441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,1024,0.017822222577200997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,256,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,2048,0.013131555583741931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,256,0.010022222167915767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,1536,0.018981332580248516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,256,0.0114720000161065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,2560,0.02628266645802392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,512,0.01110666659143236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,512,0.008362666600280339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,64,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,128,0.008016000191370646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,512,0.013361777696344586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,64,0.009688888986905416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,32,0.009700444837411245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,256,0.008012444608741337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,65536,0.4467715687221951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,128,0.009705777797434065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,65536,0.17790844705369738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,65536,0.4475590917799208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,32,0.006314666734801398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,16384,0.12560266918606228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,51200,0.35082043541802305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,128,0.006641777853171031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,16384,0.11227911048465306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,16384,0.04897422260708279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,12288,0.09607111083136664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,65536,0.4824506441752116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,128,0.01144266708029641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,51200,0.3684951199425592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,12288,0.08526755703820123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,10240,0.0609075559510125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,12288,0.09535733196470474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,51200,0.14012800322638616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,10240,0.032940444019105695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,10240,0.07200444406933255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,10240,0.08146044280793932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,7168,0.05495822098520067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,8192,0.027277333868874445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,8192,0.05964177846908569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,16384,0.12326488229963516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,8192,0.06799199846055773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,6144,0.048488001028696694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,51200,0.3729146586524115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,12288,0.03852622376547919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,5120,0.041602667835023664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,6144,0.04673066735267639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,8192,0.0643448895878262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,5120,0.039911111195882164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,4096,0.03367644548416138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,5120,0.04667199982537163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,7168,0.02492444382773505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,4096,0.0329786671532525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,3584,0.02815822097990248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,4096,0.01756266587310367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,6144,0.022094221578703985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,6144,0.053542223241594106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,3584,0.01610844499535031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,3584,0.029393778906928167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,5120,0.019860444797409903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,3072,0.014901333385043673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,2560,0.021528000632921856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,7168,0.05300355619854397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,4096,0.039649777942233615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,2560,0.023042667243215773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,2048,0.01756444407833947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,2560,0.028590222199757893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,3584,0.0366559988922543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,2048,0.012088889049159156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,2048,0.024881778491867915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,3072,0.03204177816708883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,7168,0.06064266628689236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,1536,0.016875555117925007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,1024,0.01038044442733129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,1536,0.010210666391584609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,3072,0.02476799984773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,1024,0.013331555657916598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,768,0.009373333719041612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,3072,0.02624711063173082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,1536,0.013788444300492605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,768,0.00869066682126787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,768,0.011405333048767514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,768,0.015108444624476962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,2560,0.013565333353148567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,512,0.010073777702119615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,256,0.006242666807439592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,512,0.012659555508030785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,1024,0.009022222624884712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,1024,0.017550221747822232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,128,0.0052915554907586836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,256,0.00795555528667238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,256,0.011066666907734342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,512,0.0074240002367231585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,128,0.00905955582857132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,128,0.007713778151406183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,128,0.010732444624106089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,1536,0.021474666065639917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,64,0.00904444439543618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,512,0.00832444429397583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,32,0.008976000050703684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,2048,0.01979022224744161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,65536,0.3015786541832818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,256,0.00924711094962226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,51200,0.25282133950127494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,65536,0.3687911033630371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,64,0.005244444641802046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,32,0.005551110953092575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,16384,0.0943911141819424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,51200,0.2865413294898139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,51200,0.10757955577638413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,12288,0.07309689124425252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,16384,0.037957333856158786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,16384,0.09240621990627712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,16384,0.09037422471576267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,10240,0.05926666657129923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,51200,0.2716195583343506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,12288,0.07042933172649808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,65536,0.13793067137400308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,8192,0.04975822236802843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,10240,0.025868443979157343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,10240,0.06079910861121284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,8192,0.021391999390390184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,7168,0.04439022143681844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,8192,0.050741334756215416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,65536,0.34249154726664227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,6144,0.03825155562824673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,12288,0.0702106687757704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,7168,0.019474666979577806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,7168,0.043471111191643604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,5120,0.031223111682467993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,7168,0.045666668150160045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,10240,0.05980000231001112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,6144,0.04075288772583008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,5120,0.016153777639071148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,4096,0.02537511123551263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,5120,0.03552800085809495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,4096,0.014420444766680399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,4096,0.027806222438812256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,3584,0.022825777530670166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,4096,0.030686222844653662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,3584,0.013460444079505073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,3584,0.025064888927671645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,3072,0.01962844365172916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,3584,0.028024888700909082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,5120,0.03279377685652839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,3072,0.022422222627533808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,8192,0.04887110988299052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,3072,0.025504888759719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,2560,0.01129244433508979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,12288,0.029878222280078467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,2560,0.023025777604844835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,6144,0.01796888808409373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,6144,0.03827111257447137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,1536,0.011916444533401065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,2048,0.010029333333174387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,3072,0.012425777812798819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,2560,0.017642666896184284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,1024,0.009959110783206092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,1536,0.014313777287801107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,1536,0.017681777477264404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,2560,0.0196115556690428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,1024,0.011374221907721626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,1024,0.014081777797804939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,2048,0.01690933273898231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,768,0.009870222045315636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,512,0.006614222294754452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,768,0.007842666572994655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,1536,0.008719999757077958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,512,0.009027555584907532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,256,0.005305777821275923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,512,0.0111191107167138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,512,0.007382222347789341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,768,0.008439111212889353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,128,0.004599110947714912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,256,0.007282666862010956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,256,0.008347555167145198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,256,0.01017955525053872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,64,0.0042275554604000514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,2048,0.020097777247428894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,128,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,768,0.012723555167516073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,128,0.010102222363154093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,64,0.008043555749787224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,65536,0.300780455271403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,2048,0.014170666535695394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,1024,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,65536,0.33930312262641055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,51200,0.23808712429470488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,65536,0.13033689392937556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,128,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,51200,0.26575912369622123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,65536,0.325476434495714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,16384,0.08716355429755317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,32,0.008034666379292807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,12288,0.07058133019341363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,16384,0.03693688909212748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,16384,0.08825333250893487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,12288,0.02900000082121955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,10240,0.058465778827667236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,16384,0.09280622005462646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,32,0.004597333156400257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,10240,0.057202663686540395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,10240,0.024975111087163288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,8192,0.04745955599678887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,10240,0.05995733208126492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,8192,0.020920889245139226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,8192,0.047224889198939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,7168,0.040087110466427274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,8192,0.0493182208802965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,12288,0.06938488615883721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,6144,0.03602844476699829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,7168,0.04195288817087809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,51200,0.10165066851509942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,7168,0.04500710964202881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,5120,0.02953688965903388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,6144,0.01758577757411533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,6144,0.03976710968547397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,5120,0.031700445546044245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,5120,0.015987555185953777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,4096,0.024572443630960252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,5120,0.03507733345031738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,4096,0.014355555176734924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,7168,0.019375110665957134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,3584,0.02271288964483473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,4096,0.030223998758527968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,3072,0.019666666785875957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,3584,0.013213333156373767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,6144,0.03705688979890611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,3584,0.027549333042568628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,51200,0.2544035646650526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,3072,0.012088889049159156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,3072,0.02520977788501316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,12288,0.0677128897772895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,2560,0.01919911139541202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,2048,0.015557333827018738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,2560,0.010876444478829702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,2560,0.022829333941141765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,4096,0.0270106659995185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,2048,0.009646221995353699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,2048,0.016297777493794758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,2048,0.019996444384257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,1024,0.010020444790522257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,1536,0.0138586668504609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,2560,0.0177208897140291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,1536,0.017480888300471835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,768,0.008423999779754216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,1024,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,1024,0.013939556148317126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,768,0.009688888986905416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,1536,0.012968888713253869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,768,0.012483555409643384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,3584,0.024357333779335022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,512,0.011427555647161273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,3072,0.021707556313938562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,512,0.007318221860461765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,1024,0.011071110765139261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,256,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,256,0.0070471109615431885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,128,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,256,0.010035555395815108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,512,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,64,0.004567111117972268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,128,0.007099555598364935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,128,0.007654222349325816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,1536,0.00868444475862715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,64,0.007687110867765214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,256,0.005586666779385672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,32,0.007754666937722101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,65536,0.24963821305168998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,51200,0.19637244277530244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,65536,0.12168533272213405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,65536,0.29810576968722874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,768,0.007703999678293864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,16384,0.07852266894446479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,32,0.0063457778758472866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,51200,0.09780089060465495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,512,0.008667555948098501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,65536,0.32505954636467826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,51200,0.23055378595987955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,12288,0.06052355633841621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,16384,0.03570400012864007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,51200,0.2555297745598687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,10240,0.050331556134753756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,12288,0.028082665469911363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,12288,0.059621334075927734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,8192,0.03984444340070089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,12288,0.06901955604553223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,10240,0.05112355616357592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,10240,0.05912799967659844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,8192,0.020720889170964558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,8192,0.04169422056939867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,8192,0.04946310983763801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,7168,0.018969777557584975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,6144,0.03071555495262146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,16384,0.08824711375766331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,10240,0.02454400062561035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,6144,0.03307466705640157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,6144,0.03934222128656175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,6144,0.017438222964604694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,128,0.009696000152164036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,5120,0.028759999407662287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,7168,0.044488890303505786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,4096,0.02089511189195845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,5120,0.025546666648652818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,4096,0.024255111813545227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,4096,0.013794666363133324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,3584,0.01889955500761668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,4096,0.029715554581748113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,5120,0.03450399968359206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,3584,0.02202933364444309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,3584,0.0127688886390792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,3584,0.027266666293144226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,7168,0.03489689032236735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,3072,0.01979022224744161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,3072,0.01147466649611791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,7168,0.037463999456829496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,2560,0.017326222525702584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,2560,0.010491555763615502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,2560,0.02219555609756046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,3072,0.016787555482652452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,2048,0.015003555350833468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,2048,0.009339555270142024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,2048,0.01980799933274587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,2560,0.01475822263293796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,3072,0.024877332978778418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,1024,0.008781332936551835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,1536,0.012440000143316058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,1536,0.00867377801073922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,1536,0.01662400033738878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,768,0.007408888803588019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,1024,0.007976888782448238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,2048,0.013105777402718862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,1024,0.010057777994208867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,512,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,768,0.0076657773719893555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,768,0.008987555901209513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,768,0.012473777764373355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,1536,0.01036088913679123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,512,0.007297777467303806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,512,0.01129244433508979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,16384,0.07733066876729329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,256,0.0069848886794514125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,128,0.004944889081848992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,256,0.010363555616802638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,5120,0.01573244399494595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,64,0.004430222428507275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,128,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,32,0.004612444589535396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,128,0.010022222167915767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,64,0.007343110938866933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,256,0.005514666852023866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,512,0.008153777983453538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,256,0.0075039996041191956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,65536,0.11927910645802815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,65536,0.2536719905005561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,51200,0.18927200635274252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,1024,0.013765333427323235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,65536,0.32696533203125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,51200,0.2010693285200331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,65536,0.21617421838972303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,51200,0.25439733929104275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,16384,0.0671280026435852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,12288,0.04605688982539707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,16384,0.08689510822296143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,16384,0.03374844458368089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,10240,0.04105955693456862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,12288,0.026884444885783728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,12288,0.06844000021616618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,128,0.007340444458855524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,10240,0.043980442815356784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,16384,0.06320799721611871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,10240,0.023896889554129705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,32,0.007318221860461765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,8192,0.020089776979552377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,7168,0.027431999643643696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,8192,0.048503110806147255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,7168,0.018402665853500366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,7168,0.044228444496790566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,8192,0.03307377629809909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,51200,0.0951217810312907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,10240,0.058232002788119845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,5120,0.020384889509942796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,6144,0.016879111528396606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,8192,0.03625599874390496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,6144,0.03867200016975403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,4096,0.017112000121010672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,5120,0.015284443895022074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,5120,0.03428533342149522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,12288,0.0514613323741489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,6144,0.0247999992635515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,3584,0.015832889411184523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,4096,0.013252444565296173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,4096,0.02900622288386027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,3072,0.013880000346236758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,3584,0.012158222496509552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,6144,0.028873778051800195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,3584,0.02681333323319753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,3072,0.01793688866827223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,3072,0.011389333340856763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,3072,0.024119999673631456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,5120,0.025394666526052687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,7168,0.0324888891643948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,2560,0.009750222166379293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,4096,0.021614222062958613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,2048,0.013439999686347114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,2048,0.018799111247062683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,2048,0.00904177791542477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,1536,0.008996444443861643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,2560,0.012277333272827996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,1536,0.01108088923825158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,1024,0.007241778075695038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,2560,0.01575911045074463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,2560,0.021822222405009802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,1024,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,3584,0.01969688965214623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,768,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,768,0.008354666332403818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,768,0.012079111403889127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,512,0.0053137776752312975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,1536,0.016198222835858662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,512,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,256,0.004579555657174853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,512,0.010738666686746808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,1024,0.00776711106300354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,1024,0.01311555587583118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,128,0.004238222208287981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,256,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,256,0.007048889166778988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,256,0.009701333112186855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,64,0.003949333396222857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,128,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,128,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,32,0.004224888980388641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,768,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,2048,0.01074488874938753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,65536,0.17251733938852945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,512,0.007086222370465596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,51200,0.1370097796122233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,65536,0.1763155592812432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,51200,0.16639199521806505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,1536,0.008374222450786168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,16384,0.054494222005208336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,64,0.006950221955776215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,51200,0.14035822285546198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,16384,0.02276977731121911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,12288,0.04073244333267212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,12288,0.04358933369318644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,128,0.009757333331637913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,51200,0.061648892031775586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,32,0.006959111326270633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,10240,0.03470844361517165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,12288,0.041913777589797974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,65536,0.21083466211954752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,8192,0.027684445182482403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,10240,0.01687733332316081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,10240,0.03611377875010172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,8192,0.030999110804663763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,7168,0.02420622275935279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,12288,0.018981332580248516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,7168,0.028018666638268366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,65536,0.07547911008199056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,6144,0.020994666549894545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,16384,0.05178222060203552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,6144,0.025107555919223364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,6144,0.025456888808144465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,5120,0.017567111386193168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,5120,0.022032888399230108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,16384,0.05637600024541219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,7168,0.013444444371594323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,10240,0.03744533326890733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,4096,0.015182221929232279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,5120,0.01108088923825158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,5120,0.022487110561794702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,4096,0.009747555686367882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,3584,0.013780444032616086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,8192,0.03129600154028998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,4096,0.0199955552816391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,6144,0.01236533373594284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,3584,0.00943022221326828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,3584,0.017264000243610807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,3584,0.01874755488501655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,2560,0.01111288865407308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,3072,0.015509333875444202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,7168,0.028930667373869155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,2560,0.013549333645237817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,2560,0.008346666892369589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,2560,0.015488000379668342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,4096,0.018637332651350234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,3072,0.012300444145997366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,2048,0.01165244479974111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,2048,0.013680000272062091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,3072,0.00871377769443724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,1536,0.00998133338159985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,1024,0.00598933340774642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,3072,0.016914667354689706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,1536,0.007341333561473423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,1536,0.01202933324707879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,2048,0.00980799976322386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,1024,0.006979555719428592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,1024,0.010450666977299584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,2048,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,768,0.007631111476156447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,512,0.004791111167934206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,768,0.006680000159475539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,768,0.010029333333174387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,512,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,512,0.007087111473083496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,256,0.0042151109211974675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,512,0.009359999663299983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,8192,0.014456000592973499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,256,0.00630044440428416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,128,0.003988444391224119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,256,0.008716444174448649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,1536,0.008042666647169325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,64,0.0038853333228164246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,128,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,128,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,32,0.00388355553150177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,64,0.006404444575309753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,32,0.006635555376609166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,768,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,256,0.006646222124497096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,65536,0.20729333824581572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,51200,0.11827199988894993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,65536,0.17229510678185356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,128,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,16384,0.04536799920929802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,51200,0.058556444115108915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,65536,0.19251644611358643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,12288,0.037668443388409086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,16384,0.022153778208626643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,51200,0.13572443856133357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,16384,0.0503786669837104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,10240,0.029648890097935993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,12288,0.04102222124735514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,10240,0.03619822197490268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,51200,0.16229600376553005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,10240,0.01642400026321411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,16384,0.05475911166932848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,1024,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,8192,0.030249777767393324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,12288,0.018294221825069852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,12288,0.042281776666641235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,8192,0.014256889621416727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,8192,0.030638221237394545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,6144,0.018757333358128864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,7168,0.013111111190583972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,7168,0.027282666828897264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,7168,0.028007109959920246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,5120,0.016442666451136272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,6144,0.011750222080283694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,8192,0.024813334147135418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,6144,0.02588622272014618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,5120,0.01072799993885888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,5120,0.021368000242445204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,5120,0.022202667262819078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,7168,0.02033333314789666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,3584,0.0124195557501581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,65536,0.07252889209323458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,4096,0.019766221443812054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,3584,0.009141333401203156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,10240,0.036193778117497764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,6144,0.024587555064095393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,4096,0.013801777528391944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,3072,0.014868444866604276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,2560,0.010395555860466426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,4096,0.009682666924264696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,2560,0.012786666552225748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,3584,0.016528889536857605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,2560,0.008117333054542542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,3584,0.01861422260602315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,2048,0.011110222174061669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,2048,0.013163555827405719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,3072,0.008729777402347988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,1536,0.007632888853549957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,4096,0.018346667289733887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,1536,0.007318221860461765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,1024,0.006239111224810283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,3072,0.01145511120557785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,1024,0.008347555167145198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,1024,0.010430222584141625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,768,0.005623111294375525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,2048,0.00777422222826216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,1536,0.009703111317422655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,1536,0.01202933324707879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,512,0.004905777672926585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,2560,0.014955555399258932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,512,0.0069413334131240845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,3072,0.01706133286158244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,256,0.00424888895617591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,512,0.006562666760550604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,768,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,768,0.009858667022652095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,256,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,128,0.003928889003064898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,128,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,128,0.008566222257084316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,512,0.009022222624884712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,64,0.00370399985048506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,128,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,32,0.003900444342030419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,256,0.006641777853171031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,768,0.006637333167923822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,65536,0.1681520011689928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,65536,0.17081956068674722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,65536,0.17013955116271973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,65536,0.07139644357893202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,64,0.006452444526884291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,2048,0.008998221821255153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,51200,0.1329395506117079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,16384,0.04272622201177809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,51200,0.13451377550760904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,1024,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,16384,0.045928001403808594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,16384,0.02180800007449256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,12288,0.032484445306989886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,51200,0.1304897732204861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,16384,0.049991110960642494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,10240,0.02751022246148851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,12288,0.01824977828396691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,12288,0.035465776920318604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,8192,0.02218311031659444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,10240,0.030549334155188665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,51200,0.05739555756251017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,10240,0.03503288825352987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,7168,0.019606222709019978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,8192,0.02590844366285536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,8192,0.029671112696329754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,7168,0.023240889112154644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,6144,0.01721333298418257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,32,0.0063928887248039246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,12288,0.04104177819357978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,10240,0.016442666451136272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,5120,0.014936000108718872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,6144,0.011644444531864591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,6144,0.025404444999165002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,8192,0.013800889253616333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,4096,0.013052444491121503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,5120,0.018610666195551556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,5120,0.022310222188631695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,7168,0.012897777888509961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,3584,0.011773332953453064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,4096,0.01555999947918786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,4096,0.009687999884287516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,4096,0.019567999574873183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,3072,0.010628444453080496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,3584,0.014159111513031853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,6144,0.02109866672092014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,3584,0.017858665850427415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,2560,0.009575111170609793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,3072,0.00868711123863856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,3072,0.016178665889634024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,5120,0.010757333702511258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,2560,0.011356444822417365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,2560,0.008045333127180735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,2560,0.014677332507239448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,7168,0.027442667219373915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,1536,0.0070062221752272705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,2048,0.007679999702506595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,3584,0.009016888837019602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,2048,0.013172444370057849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,1024,0.00591822216908137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,1536,0.00868888861603207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,1536,0.01187999960449007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,1024,0.007547555698288812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,1024,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,768,0.005186666631036335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,2048,0.008310221963458592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,768,0.006675555474228329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,768,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,256,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,768,0.00980266680320104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,512,0.00630044440428416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,1536,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,256,0.004344888859324985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,512,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,256,0.006237333433495627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,256,0.0061039999127388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,256,0.008401778009202745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,512,0.004601777841647466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,2048,0.010074666804737514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,64,0.0038737778862317405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,128,0.006263111200597551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,32,0.0038657776183552216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,512,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,3072,0.01290844463639789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,65536,0.14668356047736272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,32,0.006144000010357962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,1024,0.010398222340477837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,51200,0.11632800102233887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,65536,0.16892266273498535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,128,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,51200,0.12934311230977377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,16384,0.03835555579927232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,51200,0.05619111326005724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,51200,0.13395110766092935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,128,0.0038737778862317405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,65536,0.16647199789683023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,16384,0.04917066627078586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,12288,0.02934488985273573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,128,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,12288,0.017847999930381775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,10240,0.024323556158277724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,12288,0.04035733474625482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,64,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,10240,0.015848888291252982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,10240,0.03033422099219428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,8192,0.020000000794728596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,10240,0.03512622250450982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,65536,0.06966044505437215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,16384,0.021535999245113794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,7168,0.017549332645204332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,8192,0.02533422244919671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,8192,0.029439111550649006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,6144,0.015559110376569959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,7168,0.022767111659049988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,7168,0.027662222584088642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,5120,0.013589333328935834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,6144,0.011069333387745751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,6144,0.020619556307792664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,6144,0.024098666177855596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,4096,0.011750222080283694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,5120,0.010045333041085137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,5120,0.022118222382333543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,8192,0.013473778135246701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,4096,0.015312888556056552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,3584,0.011029333704047732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,4096,0.01867999964290195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,12288,0.03472000029351976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,7168,0.012155555188655853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,3072,0.010221333139472539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,3584,0.01386666629049513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,3584,0.01739377776781718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,2560,0.00888177752494812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,3072,0.008376888930797577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,3072,0.012454222473833295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,16384,0.044509334696663745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,3072,0.01590666671593984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,5120,0.01830044388771057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,2560,0.008016000191370646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,2560,0.010968888799349466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,1536,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,2048,0.007648000286685095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,4096,0.009358222285906473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,2048,0.013291555146376291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,1536,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,1536,0.00868622213602066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,3584,0.008972444468074376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,2048,0.007689778175618913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,768,0.004564444637960858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,2560,0.014561777313550314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,2048,0.00978488889005449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,768,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,1024,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,512,0.004212444441186057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,1536,0.01181333346499337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,512,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,256,0.0038693332009845306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,512,0.009372444616423713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,1024,0.007641777396202087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,256,0.006096000058783426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,256,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,256,0.008399110701349046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,768,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,128,0.00870488915178511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,128,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,64,0.0035626664757728577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,32,0.0035582222044467926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,1024,0.006694222076071634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,64,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,32,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,65536,0.12791200478871664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,768,0.009340444372759925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,65536,0.16098666191101074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,512,0.0063884444534778595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,65536,0.06843199994828966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,1024,0.010438222024175854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,51200,0.05527199970351326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,51200,0.10008711285061306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,128,0.0035537779331207275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,51200,0.12798133161332872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,12288,0.02680088910791609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,16384,0.021016000045670405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,16384,0.047022223472595215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,128,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,10240,0.022679110368092854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,51200,0.0806079970465766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,12288,0.03751022285885281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,10240,0.02419822249147627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,65536,0.10797422462039524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,16384,0.034875555170906916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,8192,0.0203840004073249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,8192,0.01312800000111262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,12288,0.01737777723206414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,7168,0.0174319992462794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,8192,0.027727110518349543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,10240,0.015438222222858004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,7168,0.018904000520706177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,8192,0.0192248887485928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,6144,0.015187554889255099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,7168,0.011689778003427716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,6144,0.011023999916182624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,5120,0.013153778182135688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,6144,0.02295733326011234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,12288,0.027801778581407335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,5120,0.010048888623714447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,5120,0.014588443769348992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,4096,0.011426666544543372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,5120,0.02012800011369917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,10240,0.03268355462286208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,7168,0.02557777696185642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,4096,0.009382222261693742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,4096,0.012129777835475074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,4096,0.01738133364253574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,3072,0.009372444616423713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,3584,0.008735111190213097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,3584,0.011424000064531961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,3584,0.01650044487582313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,2560,0.008085333638721043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,3072,0.008445333275530074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,3072,0.010729778144094678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,3072,0.014870221416155497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,2048,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,16384,0.03478666808870103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,2560,0.0136835558546914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,3584,0.010031110710567897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,1536,0.005981333139869902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,2048,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,2048,0.012699555191728802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,6144,0.01687022215790219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,1024,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,1536,0.007297777467303806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,1536,0.011409777734014722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,768,0.004569777597983678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,1024,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,1024,0.00980266680320104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,2560,0.008085333638721043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,512,0.0042168887125121225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,768,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,768,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,768,0.009383111364311641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,2048,0.008599110775523717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,512,0.006302222195598815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,512,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,512,0.00868000007337994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,1536,0.007661333514584436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,256,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,256,0.008375110725561777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,1024,0.00664533343580034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,64,0.0036524443162812125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,128,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,32,0.0038986665507157645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,64,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,256,0.0038986665507157645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,65536,0.08799022436141968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,32,0.005614222337802251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,256,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,65536,0.06741155518425836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,65536,0.12150843938191731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,65536,0.16034400463104248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,128,0.005986666513813867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,51200,0.09537333250045776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,51200,0.05372977919048733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,51200,0.1268506712383694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,16384,0.020639111598332722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,16384,0.04659733176231384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,128,0.003744889050722122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,51200,0.07057688633600871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,12288,0.026991999811596338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,12288,0.01703555550840166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,10240,0.017288888494173687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,12288,0.0365920000606113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,128,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,8192,0.014716444744004143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,10240,0.014474666780895658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,10240,0.024064888556798298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,10240,0.0321706665886773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,16384,0.03372177812788222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,7168,0.013432888521088494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,12288,0.020263110597928364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,8192,0.027273777458402846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,6144,0.012181333369678922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,7168,0.011476444701353708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,7168,0.017872888180944655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,7168,0.024470221665170457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,5120,0.010760000182522668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,6144,0.01657511128319634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,6144,0.02235288918018341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,5120,0.014420444766680399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,4096,0.00942666663063897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,5120,0.019501333435376484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,8192,0.019888889458444383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,4096,0.00906666699382994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,3584,0.008773333496517604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,4096,0.017246221502621967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,6144,0.01075466639465756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,16384,0.028218666712443035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,3584,0.011343999869293638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,3072,0.007815111014578078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,3584,0.0159288893143336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,5120,0.010034666293197209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,8192,0.012311110893885294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,3072,0.008300444318188561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,2560,0.007365333537260692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,3072,0.01459111107720269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,4096,0.01222222215599484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,2048,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,2560,0.007962666451931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,2560,0.009384888741705153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,2560,0.013859555953078799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,2560,0.009359999663299983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,2048,0.008358221915033128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,2048,0.012201777762836881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,1536,0.007681777907742395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,1024,0.004559111264016894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,1536,0.01107022249036365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,1024,0.006641777853171031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,1024,0.009704888694816166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,768,0.004575110971927643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,1536,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,3584,0.008654221892356873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,768,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,1536,0.007079111205206976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,512,0.003882666842805015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,3072,0.010432889064153036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,256,0.003543111185232798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,512,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,512,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,512,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,128,0.0032702222880389956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,256,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,1024,0.006918222126033571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,2048,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,64,0.003212444484233856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,128,0.005961777849329843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,32,0.003540444291300244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,128,0.00832622249921163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,768,0.006610666712125142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,768,0.00908088849650489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,64,0.005607999861240387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,32,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,65536,0.0654346677992079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,65536,0.08110400040944417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,65536,0.15916621685028076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,256,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,51200,0.06400711006588407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,51200,0.052024887667761914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,128,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,51200,0.1258817778693305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,12288,0.019175999694400363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,16384,0.01965688831276364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,65536,0.06160710917578804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,16384,0.04606489009327359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,10240,0.016788444585270353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,12288,0.019318222999572754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,12288,0.03643555442492167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,256,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,10240,0.01686311099264357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,16384,0.01940711173746321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,8192,0.014290667242474027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,16384,0.023802666200531855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,8192,0.027110222313139174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,8192,0.012152888708644442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,12288,0.015657777587572735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,51200,0.04766133427619934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,6144,0.011389333340856763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,7168,0.013159111142158508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,7168,0.024438222249348957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,8192,0.014051554931534661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,5120,0.010066666536860997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,6144,0.012094222009181976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,6144,0.01072622256146537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,6144,0.021935999393463135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,4096,0.008433777425024245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,5120,0.009735110733244155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,5120,0.010880000061459012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,7168,0.013097777962684631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,5120,0.019256000717480976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,3584,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,4096,0.009727111293209923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,4096,0.01703733371363746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,3072,0.00701333334048589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,3584,0.00869333330127928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,10240,0.014147554834683737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,3584,0.015915556086434256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,10240,0.03176000052028232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,2560,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,3072,0.008433777425024245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,3072,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,3072,0.014535999960369535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,2560,0.008039111064540016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,2560,0.007685333490371704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,2560,0.013812444276279874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,2048,0.0070053330726093715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,4096,0.009003555609120263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,1536,0.004982222285535601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,7168,0.011424889167149862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,3584,0.00925866680012809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,1536,0.006641777853171031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,1536,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,2048,0.005636444522274866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,1024,0.00599644457300504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,768,0.003887999802827835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,1024,0.009696000152164036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,2048,0.007680888805124495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,2048,0.012133333418104382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,768,0.005718222094906702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,512,0.003560888684458203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,768,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,1536,0.011090666883521609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,512,0.006329777754015393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,512,0.0052977779673205475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,512,0.009073778159088558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,256,0.0052728888889153796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,128,0.0031831111345026228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,256,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,128,0.005018666800525454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,768,0.009357333183288574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,128,0.00600000015563435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,32,0.0028782222006056043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,256,0.005983110931184556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,64,0.004980444494220945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,32,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,1024,0.004398222184843487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,128,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,65536,0.07946222358279757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,768,51200,0.0383200016286638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,65536,0.0640373362435235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,65536,0.15863733821445042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,768,16384,0.0169804443915685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,51200,0.05070133341683281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,51200,0.06261955367194282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,1024,0.006642666541867786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,51200,0.12575644916958278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,768,12288,0.01571733256181081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,16384,0.019104889697498746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,16384,0.04561688833766513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,768,10240,0.014871110518773397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,12288,0.036095112562179565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,64,0.0034293333689371743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,10240,0.016544888416926067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,768,8192,0.013578666581047906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,10240,0.03140711122088962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,768,65536,0.046839111381106906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,8192,0.014142221874660917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,768,7168,0.012788444757461548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,8192,0.026086222794320848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,12288,0.01869066721863217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,7168,0.011408000356621213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,768,6144,0.01128000020980835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,7168,0.023896889554129705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,10240,0.014160888062583076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,6144,0.011815999945004782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,6144,0.010712888505723743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,16384,0.02334844403796726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,8192,0.012192000117566852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,5120,0.01071466671095954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,5120,0.009762666291660732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,7168,0.012833777401182385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,256,0.003536888708670934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,4096,0.009637333452701569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,768,3584,0.007636444436179266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,4096,0.016875555117925007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,768,5120,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,3584,0.0090844440791342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,12288,0.015527110960748462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,768,3072,0.0070053330726093715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,768,4096,0.008300444318188561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,3072,0.008366222182909647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,3072,0.008340444829728868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,768,2560,0.006303999986913469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,4096,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,2560,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,6144,0.021712000171343487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,2560,0.013593778014183044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,3584,0.00869333330127928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,3584,0.015847111741701763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,2048,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,5120,0.019135110908084445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,2048,0.012094222009181976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,3072,0.014497778481907315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,1536,0.006614222294754452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,768,1024,0.004576000074545543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,1536,0.011058666639857821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,2560,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,1024,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,768,768,0.0042071110672420925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,1024,0.009697777529557547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,768,0.005622222191757626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,768,0.009121778110663096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,1536,0.00700800038046307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,512,0.0052764444715446895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,512,0.006270222365856171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,768,2048,0.005629333357016246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,512,0.00870222184393141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,768,0.0063493334584765965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,768,512,0.0038986665507157645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,256,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,768,128,0.00319377767542998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,768,256,0.003547555456558863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,768,1536,0.004970666848950916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,256,0.008375999828179678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,768,64,0.0031893334041039148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,128,0.006100444330109491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,768,32,0.003185777821474605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,64,0.004975111120276981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,32,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,1024,0.006008000009589725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,65536,0.07795377572377522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,128,0.007985778152942657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,65536,0.15819022390577528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,65536,0.061781333552466504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,51200,0.06144444147745768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,512,65536,0.034840001000298396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,512,16384,0.014537778165605334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,51200,0.048178666167789035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,16384,0.023022222850057814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,256,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,512,12288,0.012488889197508493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,512,51200,0.02734400000837114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,12288,0.018228444788191054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,12288,0.015392000476519266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,512,10240,0.014148443937301636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,12288,0.03604711095492045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,51200,0.1256355577045017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,10240,0.015969778100649517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,512,8192,0.013286222186353473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,10240,0.030914667579862807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,2048,0.007385777930418651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,512,7168,0.011770666473441653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,8192,0.025867554876539443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,7168,0.012773333324326409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,7168,0.011417778001891242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,512,6144,0.010587555666764578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,16384,0.01825777855184343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,6144,0.011751111182901593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,8192,0.012084444363911947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,512,5120,0.009353777600659264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,6144,0.010406222608354356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,6144,0.021560000048743352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,5120,0.010729778144094678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,512,4096,0.00798044436507755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,5120,0.01926577753490872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,10240,0.013887999786270989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,4096,0.009010666774378883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,512,3584,0.007445333732499017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,4096,0.016568000117937725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,16384,0.045203556617101036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,3584,0.00868622213602066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,3584,0.009006222089131674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,512,3072,0.0069573331210348345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,3584,0.01566755606068505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,5120,0.009720889230569204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,3072,0.008076444268226624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,512,2560,0.005964444329341252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,3072,0.014194667339324951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,4096,0.00941422250535753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,2560,0.007654222349325816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,512,2048,0.005591111050711737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,2560,0.013472889032628802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,2048,0.00741155528359943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,512,1536,0.004613333278232151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,2048,0.012070222033394707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,7168,0.023762666516833838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,3072,0.008385777473449707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,1536,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,512,1024,0.004254222330119875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,1536,0.011064888702498542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,2560,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,1024,0.006678222368160884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,512,768,0.0039057777159743835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,1024,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,1024,0.009746666583749983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,512,512,0.003583111282851961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,768,0.006443555570311016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,768,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,768,0.008745777938101027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,512,256,0.0032186667538351486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,512,0.00565244464410676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,128,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,1536,0.006980444822046492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,512,128,0.0029146667155954572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,256,0.005978666659858491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,256,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,8192,0.013793778088357715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,512,64,0.0028942223224374983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,128,0.007974222302436829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,512,32,0.0028977776981062363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,64,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,512,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,256,65536,0.02686133318477207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,512,0.008723555339707268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,256,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,65536,0.07815555731455485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,65536,0.15831644005245632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,128,0.005984888722499211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,51200,0.059715555773841016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,51200,0.047000000874201454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,51200,0.12508532736036512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,65536,0.05951999955707126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,16384,0.02202755543920729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,128,0.005211555709441503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,256,12288,0.01313244468635983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,16384,0.018231999542978074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,12288,0.017961778574519686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,32,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,12288,0.034938666555616595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,256,10240,0.011634666886594562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,256,51200,0.020582222276263766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,10240,0.015863110621770222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,10240,0.013815110756291283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,256,8192,0.010029333333174387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,2048,0.0069919998447100324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,16384,0.044863998889923096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,256,7168,0.009724444813198512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,8192,0.013551111022631327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,8192,0.02574666672282749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,256,16384,0.011415999796655444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,12288,0.015167999598715039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,256,6144,0.009001777403884465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,7168,0.023345778385798138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,256,5120,0.008386666576067606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,6144,0.010392000277837118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,6144,0.011506666739781698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,10240,0.0303422212600708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,6144,0.021369778447681006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,5120,0.009716444545321995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,5120,0.010418666733635796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,8192,0.01182577759027481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,256,3584,0.007207999626795451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,4096,0.00904799997806549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,4096,0.016576889488432143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,3584,0.00866844422287411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,3584,0.015840889679061044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,256,4096,0.007669332954618666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,3072,0.008274666965007782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,5120,0.018583999739752877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,4096,0.009359111388524374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,3072,0.008258666429254744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,7168,0.012491555677519904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,2560,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,2560,0.008000888758235509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,256,2048,0.005362666729423735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,2560,0.01350400017367469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,2048,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,2048,0.0073315559162033936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,2048,0.012090666426552666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,256,2560,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,3072,0.014155555102560254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,1536,0.006542222367392645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,3584,0.00868888861603207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,256,3072,0.006606222026877933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,1024,0.005979555762476391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,1024,0.006607999818192587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,1024,0.009311999711725447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,256,1536,0.004855999930037392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,768,0.005619555711746216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,256,512,0.0035377778112888336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,768,0.009026666482289633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,256,1024,0.003987555702527364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,512,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,256,256,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,512,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,256,768,0.003569777641031477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,256,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,256,128,0.0028453332682450614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,256,0.008013333711359236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,768,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,128,0.004961777892377642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,128,0.006004444426960415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,512,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,256,32,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,64,0.005209777918126848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,128,65536,0.017679111825095285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,256,0.005976000179847081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,65536,0.07606755362616645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,7168,0.011056000159846412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,256,64,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,65536,0.05635644329918755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,51200,0.0584684411684672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,51200,0.12530399693383112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,128,16384,0.010054222411579555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,1536,0.00701422244310379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,16384,0.022035555707083807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,65536,0.1584871080186632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,128,12288,0.008730666504965888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,16384,0.01824711097611321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,16384,0.04415466719203525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,12288,0.017948443690935772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,128,10240,0.009272889130645329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,12288,0.015152000718646579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,10240,0.015792888071801927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,1536,0.010823110739390055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,10240,0.03019555409749349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,128,8192,0.008343111309740279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,10240,0.013830222189426422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,8192,0.01348533315791024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,8192,0.025735111700163946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,128,7168,0.00795555528667238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,12288,0.0346959994898902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,7168,0.012463111016485425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,7168,0.023782221807373896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,7168,0.011075555450386472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,6144,0.011391999820868174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,32,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,51200,0.04537599947717455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,128,5120,0.00811111099190182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,6144,0.010398222340477837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,6144,0.021309332715140447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,128,4096,0.007626666790909237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,5120,0.009750222166379293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,5120,0.010321777727868822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,128,6144,0.007696000238259633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,5120,0.018998222218619455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,4096,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,4096,0.01666755477587382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,3584,0.008678221868144141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,128,3072,0.006591111007663939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,3584,0.015524443652894763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,3072,0.008280889027648503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,128,51200,0.015145777000321282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,3072,0.014145778285132514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,128,2560,0.006119110931952794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,3072,0.008006222546100616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,4096,0.009378666679064432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,2560,0.007700444095664554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,2560,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,2560,0.013194666968451606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,3584,0.008555555509196388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,128,1536,0.004586666822433472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,2048,0.007325333025720384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,2048,0.011867555479208628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,1536,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,1536,0.0063777777055899305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,128,1024,0.003882666842805015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,1536,0.011051555474599203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,128,3584,0.006959111326270633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,128,768,0.0035333335399627686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,1024,0.006603555546866522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,128,2048,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,1024,0.009691555466916826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,8192,0.012030222349696688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,128,512,0.003544888976547453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,768,0.00565244464410676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,2048,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,128,256,0.0029271110478374693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,512,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,512,0.0086666668454806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,128,0.00795911086930169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,128,128,0.0028808888875775864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,256,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,256,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,128,0.004939555707905027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,128,64,0.0025191110455327565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,128,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,128,32,0.002524444419476721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,64,0.004933333231343163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,64,65536,0.01629066632853614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,256,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,64,51200,0.013832000394662222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,1024,0.00600622221827507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,65536,0.07448177867465548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,128,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,51200,0.05749511056476169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,64,12288,0.008380444513426887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,768,0.009344889058007134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,16384,0.02167111138502757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,64,8192,0.008162666526105668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,12288,0.01771377854877048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,64,7168,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,32,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,8192,0.013423999978436364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,7168,0.012427555190192329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,64,5120,0.00798044436507755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,768,0.006452444526884291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,64,4096,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,64,10240,0.009073778159088558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,64,3584,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,5120,0.010390222072601318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,64,3072,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,4096,0.009328000247478485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,10240,0.015636444091796875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,3584,0.008589333130253686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,64,2048,0.0061039999127388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,3072,0.008212444682916006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,64,1536,0.004543111142185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,2560,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,64,1024,0.0038808890514903595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,2048,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,64,768,0.003530666646030214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,1536,0.006313777632183499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,64,512,0.003195555466744635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,1024,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,64,256,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,768,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,64,128,0.002536000063021978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,512,0.0052986666560173035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,64,6144,0.007303111255168915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,256,0.005018666800525454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,64,16384,0.009739555418491364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,128,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,64,32,0.0028862222615215513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,6144,0.011422221859296164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,32,65536,0.028143111202451918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,64,0.004947555561860402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,32,51200,0.023065777288542852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,32,16384,0.00904444439543618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,65536,0.0743173360824585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,51200,0.05817333194944593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,64,2560,0.0064497776329517365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,16384,0.02180266711446974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,32,10240,0.007646222081449296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,32,8192,0.008316444853941599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,64,64,0.0028168888141711554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,12288,0.017607111069891188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,10240,0.015543111496501498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,32,6144,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,512,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,32,5120,0.006990222467316522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,32,12288,0.008567999634477828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,32,4096,0.00759733302725686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,5120,0.010466666685210334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,32,3584,0.0070186663005087115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,4096,0.009033777647548253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,32,3072,0.006879999819729064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,32,2560,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,8192,0.01348888874053955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,32,2048,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,3072,0.008023999631404877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,32,1536,0.004271111140648524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,6144,0.011390222443474663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,32,1024,0.003908444609906939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,2048,0.006671111202902264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,32,768,0.0035342222286595237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,1536,0.006309333360857434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,32,512,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,1024,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,32,256,0.0028888889484935333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,3584,0.008902221918106079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,32,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,256,0.005165333135260476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,32,64,0.0028888889484935333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,2560,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,32,32,0.002560888934466574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,128,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,64,0.004976888911591636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,32,7168,0.008012444608741337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,32,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,768,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,512,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,16384,0.4181395636664496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,512,32,128,0.002834666727317704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,12288,0.31215111414591473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,51200,1.3193066914876301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,16384,0.21808266639709473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,12288,0.3063146538204617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,51200,1.3696159786648219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,7168,0.012468444804350534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,10240,0.13659021589491102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,16384,0.4220124350653754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,8192,0.11050933599472046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,8192,0.20793867111206055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,7168,0.18224088350931802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,7168,0.0970568921830919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,10240,0.28555909792582196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,10240,0.26947376463148326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,6144,0.17335822847154406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,5120,0.14765066570705837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,6144,0.08533600303861831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,6144,0.1578542259004381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,4096,0.12709866629706487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,7168,0.20231733057234022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,5120,0.07323200172848172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,3584,0.10716266764534844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,4096,0.05989155504438612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,3072,0.0986408856179979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,3584,0.05914577510621813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,12288,0.1782524453269111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,2560,0.08282666736178927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,3072,0.047332445780436196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,3072,0.08237955305311415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,2048,0.06921688715616862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,2560,0.04136711027887132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,2560,0.07118578089608087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,1536,0.053103999959097974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,2048,0.0354364448123508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,4096,0.10839111275143093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,2048,0.05957333246866862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,1024,0.03727822171317206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,1536,0.028314666615592107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,1536,0.04714666803677877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,768,0.03130311104986403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,1024,0.022170666191313002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,1024,0.03626666797531976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,768,0.020152888364262052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,768,0.02971022327740987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,512,0.018552889426549275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,5120,0.131658673286438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,256,0.02240000002914005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,51200,0.6693138016594781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,256,0.0170151111152437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,128,0.011898666620254517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,512,0.022994667291641235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,64,0.010851555400424533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,128,0.016770665844281513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,32,0.013448889056841532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,512,0.025045333637131587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,64,0.020986666282018025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,32,0.020999999509917364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,8192,0.2313715616861979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,128,0.021311110920376245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,256,0.01423377792040507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,65536,1.3704631593492296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,51200,1.0609750747680664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,3584,0.09775822030173408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,51200,1.0881688859727647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,65536,0.6854106585184733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,12288,0.25620532035827637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,16384,0.3278240097893609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,10240,0.2072897752126058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,12288,0.24523022439744738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,12288,0.13361956013573542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,16384,0.17251377635531953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,16384,0.33564445707533097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,10240,0.11027466588550144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,8192,0.16712533103095162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,8192,0.08932177888022529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,6144,0.14698133203718397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,7168,0.08929955297046238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,7168,0.14517866240607366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,51200,0.5275537702772353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,5120,0.11066577831904094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,6144,0.07667466666963366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,10240,0.20231200589074028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,65536,1.3397217856513128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,5120,0.05818133221732246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,3584,0.08079644706514147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,4096,0.047621332936816745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,4096,0.086081776354048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,3072,0.07052444749408297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,3584,0.04917866653866238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,3584,0.07574133078257243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,2560,0.059453335073259145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,3072,0.03730222251680162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,6144,0.1293840010960897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,3072,0.06607377529144287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,8192,0.1710240046183268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,2048,0.0524524450302124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,7168,0.15325600571102566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,2048,0.03126666612095303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,2048,0.048102223210864596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,1024,0.029469334416919287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,1024,0.029658665259679157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,4096,0.09925599892934163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,768,0.023844444089465674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,1024,0.0188071115149392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,5120,0.10446311367882623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,2560,0.05705866548750135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,512,0.01757244434621599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,768,0.016961778203646343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,768,0.02409422232045068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,512,0.0158897770775689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,1536,0.03861777649985419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,512,0.02011022302839491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,128,0.010388444695207808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,256,0.018591110905011494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,128,0.014310222533014087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,2560,0.03325422273741828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,32,0.008840888738632202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,1536,0.041570666763517596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,32,0.017595556047227647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,1536,0.023226666781637404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,65536,0.4012453291151259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,256,0.014579556054539151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,64,0.008249777886602614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,128,0.017539555827776592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,65536,0.23340445094638398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,65536,0.4428328937954373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,16384,0.10506755775875515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,64,0.017533333765135873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,51200,0.3449724515279134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,51200,0.18570044305589464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,16384,0.06172977553473579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,256,0.012784000072214337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,16384,0.11311733722686768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,12288,0.048360887500974864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,12288,0.08430666393703884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,10240,0.04154310954941644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,10240,0.07169333431455824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,51200,0.31763021151224774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,8192,0.0343795551194085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,8192,0.05887822310129801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,6144,0.049154665735032826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,7168,0.030826667944590252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,5120,0.04114222195413377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,6144,0.02732177740997738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,6144,0.045604444212383695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,10240,0.07788355482949151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,5120,0.02387199964788225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,4096,0.03382577829890781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,5120,0.0390035543176863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,3584,0.03042222062746684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,4096,0.020833777056799993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,4096,0.03256533212131924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,3072,0.026734222968419392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,3584,0.01886311173439026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,3584,0.029098666376537744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,2560,0.022721777359644573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,12288,0.07914400100708008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,3072,0.025855110751258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,2048,0.018941332896550495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,2560,0.017263111140992906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,2560,0.022523555490705702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,1536,0.014941333068741692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,2048,0.014120888378885059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,7168,0.057238221168518066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,1024,0.011760888828171624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,1536,0.01257155587275823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,1536,0.016301333904266357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,768,0.010040889183680216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,1024,0.010478222535716163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,1024,0.013237333132161034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,512,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,768,0.009718221922715506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,768,0.011415111521879831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,256,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,8192,0.06481955448786418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,512,0.009746666583749983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,128,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,2048,0.019261333677503798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,256,0.009189333352777693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,64,0.005247999810510212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,32,0.005236444373925527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,128,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,7168,0.05208177698983086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,64,0.00906044410334693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,32,0.009039999710188972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,65536,0.30180711216396755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,512,0.009121778110663096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,51200,0.23520533243815103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,65536,0.34333332379659015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,256,0.008634666601816813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,16384,0.08228800031873915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,128,0.008808000220192803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,51200,0.2655591169993083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,16384,0.05264888869391548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,16384,0.08813333511352539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,10240,0.059782220257653125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,12288,0.03886933459175958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,12288,0.06662755542331271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,8192,0.04740088846948412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,10240,0.05691822369893392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,8192,0.027504889501465693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,7168,0.03869422276814779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,7168,0.024762666887707178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,12288,0.06262666649288602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,6144,0.03375555409325494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,51200,0.15241155359480116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,3072,0.017317333155208163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,6144,0.03689600030581156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,5120,0.03189422355757819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,5120,0.019691555036438834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,4096,0.025390222668647766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,4096,0.01755288905567593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,4096,0.0262746661901474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,5120,0.03089511063363817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,3584,0.02358755634890662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,3072,0.019879110985332064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,65536,0.18270933628082275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,3072,0.021309332715140447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,2560,0.01706044375896454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,3072,0.015029332704014249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,3584,0.02257866660753886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,2048,0.014888000157144336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,2560,0.01370844410525428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,2560,0.018771555688646104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,6144,0.022276444567574397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,2048,0.012368888490729861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,2048,0.016398222910033334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,1024,0.010084444450007545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,1536,0.013829333086808523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,8192,0.04704799917009142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,1024,0.009025777379671732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,1024,0.011412444214026133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,7168,0.04193777839342753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,768,0.008667555948098501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,768,0.009694221946928237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,1536,0.012288000020715924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,512,0.008284444610277811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,512,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,1536,0.010599999792046018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,128,0.004601777841647466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,256,0.008026666939258575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,768,0.008760000268618265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,64,0.004335999902751711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,128,0.007809778054555257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,512,0.007295110987292395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,32,0.004743111216359668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,64,0.008009778128729926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,256,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,10240,0.03327288892534044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,32,0.008012444608741337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,65536,0.3109039995405409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,256,0.007770666645632849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,51200,0.2198746601740519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,16384,0.07729333639144897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,65536,0.16536800066630045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,128,0.007925333248244392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,51200,0.13271110587649876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,51200,0.22665066189236113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,12288,0.06364355484644572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,16384,0.04884533418549431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,12288,0.038078222009870745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,12288,0.05870844258202446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,8192,0.04218133290608724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,10240,0.03120000163714091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,10240,0.04986310998598734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,8192,0.02618666655487484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,8192,0.04177777634726631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,65536,0.2890426582760281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,6144,0.03316977620124817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,7168,0.025742222865422566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,7168,0.036694222026401095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,10240,0.05458044343524509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,5120,0.02878755662176344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,6144,0.02274133265018463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,4096,0.02403733299838172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,5120,0.019139554765489366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,3584,0.016141333513789706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,5120,0.027608889672491286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,4096,0.016688000824716356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,4096,0.02325422234005398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,3072,0.018710222509172227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,7168,0.037181334363089666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,3584,0.021125333176718816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,3072,0.01422933406300015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,3072,0.019167999426523846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,2048,0.014133334159851074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,2560,0.016859556237856548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,16384,0.07618933253818087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,6144,0.03233777814441257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,2048,0.011770666473441653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,2048,0.01479555500878228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,3584,0.02083288960986667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,1536,0.00998133338159985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,1536,0.012806221842765808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,3584,0.01631644368171692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,1024,0.008664888640244802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,2560,0.01609155535697937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,1024,0.009827555881606208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,512,0.006618666566080517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,768,0.00903555585278405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,2560,0.012999110751681857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,512,0.007976888782448238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,512,0.008001777860853408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,128,0.004570666700601578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,256,0.007634667058785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,1024,0.010048000348938836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,64,0.004254222330119875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,128,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,768,0.00869333330127928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,768,0.00832622249921163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,64,0.007339555356237624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,32,0.007382222347789341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,256,0.005234666582610872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,65536,0.24152088165283203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,1536,0.011907555162906647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,51200,0.2018666664759318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,65536,0.2438195546468099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,16384,0.06788000133302477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,128,0.007676444119877285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,51200,0.1013173328505622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,12288,0.0551475551393297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,16384,0.06486399968465169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,10240,0.04448799954520332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,12288,0.05003644360436333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,10240,0.024784000383483037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,10240,0.04310489032003614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,51200,0.19337333573235405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,32,0.004556444370084339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,8192,0.03580977850490146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,7168,0.03232444326082865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,8192,0.021189333664046392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,12288,0.030541333887312148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,6144,0.028263999356163874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,7168,0.021384888225131567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,5120,0.024341333243581984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,6144,0.0174346665541331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,6144,0.02773244513405694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,4096,0.01997066703107622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,8192,0.03719288773006863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,5120,0.02386133372783661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,3584,0.017877333694034152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,4096,0.014134221606784396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,4096,0.0203128887547387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,3584,0.015276443627145557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,3584,0.018597332967652213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,7168,0.031919111808141075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,3072,0.016855110724767048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,2560,0.013777777552604675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,5120,0.015844444433848064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,256,0.007648000286685095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,2048,0.012076444096035428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,2560,0.010990222295125326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,2560,0.015167999598715039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,16384,0.03566666775279575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,2048,0.009673777553770278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,2048,0.013167111410035027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,1024,0.00831911133395301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,1536,0.008734222087595198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,1536,0.011329777538776398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,768,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,1024,0.00902844468752543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,3072,0.011744000017642975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,512,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,768,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,768,0.0081386665503184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,256,0.004372444417741564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,512,0.007342221836249034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,1536,0.009867555565304225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,128,0.003919110943873723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,256,0.007111999723646376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,128,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,1024,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,64,0.004257777912749184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,32,0.004200889004601372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,65536,0.12502933873070612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,64,0.0069822221994400024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,32,0.007888889147175683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,512,0.007672000262472365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,3072,0.015560888581805758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,65536,0.1210248867670695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,16384,0.06280444728003608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,51200,0.09516177574793498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,16384,0.034882667991850115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,12288,0.05001066790686714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,65536,0.23797954453362358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,256,0.0070364442136552595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,12288,0.04292088747024536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,10240,0.04120088948143853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,51200,0.16559733284844294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,10240,0.03646666804949442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,8192,0.03554577628771464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,16384,0.05513066715664334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,7168,0.030549334155188665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,8192,0.03005511230892605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,7168,0.01869244376818339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,7168,0.026857778429985046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,6144,0.026632888449562922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,6144,0.017144888639450073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,128,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,6144,0.02407466702991062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,5120,0.015551111764378019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,8192,0.022449778185950384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,65536,0.20833244588640001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,12288,0.027339556150966223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,3584,0.016979555288950603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,4096,0.013836444252067141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,10240,0.024152888192070857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,3584,0.012802666260136498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,3584,0.01643733349111345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,4096,0.019124444988038804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,2560,0.013140444126394061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,3072,0.014927110738224454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,51200,0.19149777624342176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,2048,0.011733333269755045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,2560,0.013423110875818463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,5120,0.022431999444961548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,3072,0.015202666322390238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,2048,0.011785777906576792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,5120,0.020850666695170935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,1024,0.007640000018808577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,1536,0.010042666561073726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,2560,0.01075644459989336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,1024,0.007972444097201029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,1024,0.008356444537639618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,2048,0.009363555245929295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,768,0.007648000286685095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,3072,0.011431111229790581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,512,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,256,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,1536,0.008721777962313758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,4096,0.01789422167672051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,768,0.006264888991912206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,256,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,128,0.0069982219073507525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,64,0.004012444367011388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,64,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,32,0.004232888834344016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,128,0.004338666796684265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,65536,0.19313955307006836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,1536,0.00978488889005449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,65536,0.11686400572458903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,65536,0.1967760059568617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,512,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,51200,0.18380710813734266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,256,0.0070088886552386815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,51200,0.09313155545128717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,51200,0.15570222006903753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,32,0.006641777853171031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,512,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,16384,0.03484444485770331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,16384,0.05315733287069532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,10240,0.037018666664759316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,12288,0.026732444763183594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,12288,0.041489776637819074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,8192,0.02925155560175578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,10240,0.035473777188195124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,16384,0.05830488602320353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,8192,0.029094222519132826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,768,0.007691555553012424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,7168,0.026378666361172993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,6144,0.022462222311231825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,10240,0.023539554741647508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,6144,0.0232968893316057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,8192,0.02015199926164415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,7168,0.026570666167471144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,12288,0.04231289029121399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,5120,0.015245333313941956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,4096,0.015840000576443143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,3584,0.01440622243616316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,4096,0.013482666677898832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,4096,0.01755555636352963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,3584,0.012175111307038201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,3584,0.016009777784347534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,6144,0.017655111021465726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,3072,0.011365333365069495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,3072,0.01482488877243466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,128,0.006596444381607904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,2048,0.010119111173682744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,7168,0.01863822175396813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,2048,0.009949333137936061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,3072,0.01314222233162986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,1536,0.009005332986513773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,1536,0.00869066682126787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,2560,0.011724444727102915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,1024,0.0069875551594628235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,1024,0.007730666961934831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,2560,0.013168888787428537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,1024,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,768,0.007619555625650618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,2048,0.011436444189813403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,5120,0.019521777828534443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,1536,0.009708444277445475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,256,0.003929777691761653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,512,0.007094221810499827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,2560,0.0101724440852801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,5120,0.020432889461517334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,256,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,256,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,64,0.003543111185232798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,128,0.006993778049945831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,768,0.007382222347789341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,64,0.006331555545330048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,32,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,65536,0.20708621872795951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,768,0.006302222195598815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,65536,0.11333155632019043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,65536,0.15997421741485596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,51200,0.1567537784576416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,16384,0.05240977803866068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,32,0.0039013334446483185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,51200,0.09054222371843125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,51200,0.12439555592007107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,12288,0.04138400157292684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,16384,0.03233688738611009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,16384,0.043529778718948364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,10240,0.03413510984844632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,12288,0.02587022218439314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,12288,0.03369955552948846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,10240,0.023064000738991633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,128,0.0038897775941424896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,10240,0.02900000082121955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,7168,0.025076443950335186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,8192,0.019650666250122916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,8192,0.02425244450569153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,6144,0.02202577723397149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,7168,0.017973333597183228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,7168,0.022043555974960327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,6144,0.016349333855840895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,6144,0.019864888654814828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,4096,0.015847999188635085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,512,0.0070062221752272705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,5120,0.017304889029926725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,3584,0.014408888088332282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,8192,0.028821332587136164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,4096,0.014920888675583733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,3584,0.011983999775515662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,3584,0.013745778136783175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,512,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,3072,0.01070666644308302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,2560,0.011343110766675739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,2560,0.009771555662155151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,2048,0.009822222093741098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,2560,0.011390222443474663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,4096,0.012822222378518848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,2048,0.009190222455395592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,1536,0.008671999805503422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,2048,0.009687999884287516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,1536,0.008354666332403818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,128,0.006292444550328785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,1536,0.008393777741326226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,1024,0.007589333587222629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,5120,0.014753777119848462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,1024,0.007320889168315464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,768,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,512,0.004602666530344221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,3072,0.012901333471139273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,512,0.007136888802051544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,512,0.006335111127959357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,5120,0.018573333819707234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,256,0.00702311098575592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,768,0.005316444569163852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,768,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,128,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,64,0.003544888976547453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,128,0.006959999601046245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,32,0.003559111307064692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,256,0.004200889004601372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,3072,0.01290844463639789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,64,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,65536,0.16120178169674343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,32,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,256,0.006267555471923616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,65536,0.07073599762386747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,16384,0.0403039985232883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,65536,0.1535866657892863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,51200,0.05518933468394809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,51200,0.11973422103457981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,16384,0.04150311152140299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,10240,0.026153778036435444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,12288,0.03244444396760728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,1024,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,8192,0.021648888786633808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,10240,0.016065778003798593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,10240,0.028018666638268366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,7168,0.018903111418088276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,8192,0.014113777213626437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,8192,0.02363377809524536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,6144,0.016846223009957206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,12288,0.03044355577892727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,16384,0.021544888615608215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,5120,0.014856000741322836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,6144,0.01201244443655014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,12288,0.0181057784292433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,6144,0.019310222731696237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,4096,0.012757333616415659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,5120,0.01089422239197625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,5120,0.016861332787407767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,4096,0.00941422250535753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,4096,0.014478221535682678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,51200,0.11947288778093125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,3584,0.009123555488056606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,3584,0.013387555877367655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,2560,0.009380444056457942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,3072,0.008672888908121321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,7168,0.02144533395767212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,2048,0.00850399997499254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,2560,0.010436444646782346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,2048,0.00794400026400884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,2048,0.0094035557574696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,1536,0.006729777902364731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,1536,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,128,0.003567111161020067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,1536,0.008350222474998897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,7168,0.013139555851618448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,1024,0.006688888702127669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,3072,0.012104888757069906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,2560,0.008060444560315874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,512,0.004566222015354368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,768,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,768,0.006666666517655055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,512,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,3584,0.011381333072980246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,512,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,128,0.003552888830502828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,256,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,1024,0.005581333405441708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,64,0.003560888684458203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,128,0.005968889014588461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,768,0.005028444445795483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,32,0.003551111039188173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,1024,0.0073520003093613525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,64,0.0059164443777667145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,65536,0.13417155212826198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,32,0.005969777703285217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,51200,0.10762133200963338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,65536,0.06740355491638184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,65536,0.11779556009504531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,16384,0.04196444484922621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,51200,0.053820444477929004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,51200,0.09319111373689438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,16384,0.021091555555661518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,128,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,10240,0.024711999628278945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,12288,0.01758399936887953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,12288,0.0254737784465154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,10240,0.022050667140218947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,10240,0.01589777734544542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,3072,0.01071111112833023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,12288,0.029866665601730347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,7168,0.01865866614712609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,16384,0.03209244542651706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,7168,0.012930666406949362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,6144,0.01642666591538323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,256,0.0042257776690853965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,8192,0.020474667350451153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,5120,0.014116444521480136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,256,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,8192,0.018952889574898612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,8192,0.01386577801571952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,4096,0.012123555772834353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,5120,0.010319111247857412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,3584,0.011330666641394297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,4096,0.009242666264375051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,7168,0.01741333305835724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,4096,0.012119111087587146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,3072,0.01036622209681405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,3584,0.008691555923885768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,3584,0.011063111325105032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,2560,0.00903555585278405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,3072,0.00850844466023975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,6144,0.015879111157523263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,3072,0.009773333039548662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,6144,0.011432888607184092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,1536,0.007972444097201029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,2048,0.007536888950400882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,1024,0.006261333409282897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,1536,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,1536,0.007349333001507654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,768,0.005575110928879843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,1024,0.006686222222116258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,768,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,768,0.006305777778228124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,512,0.004592888885074192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,2048,0.009331555830107795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,2560,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,512,0.0063484443558586975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,2560,0.009005332986513773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,2048,0.008360000120268928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,256,0.006223111102978389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,256,0.005620444400442972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,64,0.0031866667171319327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,128,0.006056889063782162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,32,0.0032222221295038858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,64,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,32,0.005839111076460944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,256,0.0038844446341196695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,512,0.005961777849329843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,128,0.003535110917356279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,65536,0.06627911329269409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,51200,0.08230221933788724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,5120,0.013850666582584381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,16384,0.03379911184310913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,1024,0.006641777853171031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,51200,0.052928000688552856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,128,0.005617777920431561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,51200,0.09131111039055718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,10240,0.02389066749148899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,16384,0.031720889939202204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,12288,0.02532622218132019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,65536,0.11717600292629665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,8192,0.019076445036464267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,10240,0.02222844461599986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,7168,0.018208000395033095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,8192,0.01346933344999949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,8192,0.018907555275493197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,65536,0.11466666062672932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,6144,0.014984000060293408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,7168,0.01720533271630605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,12288,0.026748443643252056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,5120,0.013888888888888888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,16384,0.02069244450993008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,6144,0.015840000576443143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,12288,0.017256889078352187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,5120,0.01386666629049513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,10240,0.015617777903874716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,4096,0.011771555576059552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,3584,0.008656889200210571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,3584,0.010764444039927589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,7168,0.012375999655988483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,6144,0.011071999867757162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,3072,0.00977422214216656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,4096,0.012039110892348819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,5120,0.010024000373151567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,2048,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,2560,0.007779555188284979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,3584,0.010760889285140567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,1536,0.0063768890168931745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,2048,0.008307555483447181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,3072,0.011039111349317761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,1024,0.0052355556852287715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,3072,0.008336000144481659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,2560,0.009069333473841349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,768,0.004879999905824661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,1024,0.006664888726340399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,4096,0.009071110851234859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,768,0.006299555715587404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,512,0.004280888785918553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,768,0.006610666712125142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,512,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,256,0.003556444413132138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,1024,0.006605333338181178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,256,0.005583111196756363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,128,0.003530666646030214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,256,0.006146666490369373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,2560,0.008976889153321585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,2048,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,128,0.006025777922736273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,32,0.0032391111469931076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,1536,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,512,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,65536,0.09245422151353623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,65536,0.06476000282499525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,65536,0.11060355769263373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,51200,0.07441777653164335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,64,0.0032346666687064697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,51200,0.051819556289248996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,128,0.005550222264395819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,51200,0.08700711197323269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,12288,0.026290666725900438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,64,0.005588444570700328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,10240,0.021394666698243883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,12288,0.01715022159947289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,12288,0.024780443973011438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,8192,0.018043556147151522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,10240,0.02142222225666046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,8192,0.012899555265903473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,7168,0.016120000018013846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,16384,0.03149688906139798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,1536,0.007327111230956183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,7168,0.016879111528396606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,6144,0.014537778165605334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,32,0.005668444352017508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,5120,0.012628444366984896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,6144,0.010764444039927589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,6144,0.015263999501864115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,4096,0.01071466671095954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,5120,0.013383111192120446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,8192,0.01851111153761546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,3584,0.010296000374688042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,4096,0.009038222332795462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,16384,0.020370667179425556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,3584,0.010398222340477837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,16384,0.030921777089436848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,3072,0.00833422193924586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,2560,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,3072,0.009706666900051964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,5120,0.009694221946928237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,2048,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,2560,0.007640000018808577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,2560,0.00869955536392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,1536,0.006247111078765657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,2048,0.007368000017272101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,2048,0.00833777752187517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,1024,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,1536,0.006959999601046245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,3072,0.009383111364311641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,1536,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,1024,0.006643555644485686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,10240,0.01531377765867445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,512,0.0042160000238153665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,768,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,512,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,4096,0.01109155515829722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,256,0.0035573335157500375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,7168,0.01182666669289271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,256,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,128,0.003237333355678452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,256,0.005618666609128316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,64,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,128,0.006015111174848344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,32,0.003257777748836411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,64,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,65536,0.07244355811013116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,512,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,65536,0.06343555450439453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,65536,0.10612089104122585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,768,0.004576000074545543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,1024,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,128,0.0054551110499435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,16384,0.024912888805071514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,51200,0.08351555797788833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,12288,0.02071199980047014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,16384,0.020218667056825426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,32,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,10240,0.021547555923461914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,12288,0.016593777471118502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,12288,0.023990222149425085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,3584,0.008660444782839881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,10240,0.014732443624072604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,51200,0.06263555420769586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,8192,0.01239377756913503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,8192,0.017902221944597032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,51200,0.050260444482167564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,6144,0.014855111638704935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,7168,0.011391110718250275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,7168,0.01623911162217458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,16384,0.030016889174779255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,6144,0.010439111126793755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,6144,0.014459555347760519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,8192,0.0180711117055681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,5120,0.009736888938479954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,10240,0.021184000704023574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,7168,0.016623111234770883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,3584,0.00850133349498113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,4096,0.008822222550710043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,4096,0.011017777853541903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,3072,0.007795555724038019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,3584,0.008406221866607666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,3584,0.010040889183680216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,2560,0.006787555499209299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,3072,0.008108444511890411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,3072,0.009381333159075843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,2048,0.006598222172922558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,2560,0.007676444119877285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,2560,0.00868711123863856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,5120,0.012471111284361945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,2048,0.00740533322095871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,1024,0.004591111093759537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,768,0.006395555618736479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,1536,0.006983111302057902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,768,0.00423911131090588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,1024,0.00665511108107037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,1024,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,512,0.00387999994887246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,768,0.006004444426960415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,512,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,4096,0.011071110765139261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,512,0.005766222046481238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,2048,0.00813066628244188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,256,0.005979555762476391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,1536,0.007299555672539606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,128,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,128,0.0053297777970631914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,64,0.0029288888391521242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,768,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,64,0.005588444570700328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,1536,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,65536,0.059897780418395996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,128,0.0032142222755485107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,256,0.005640888793600931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,51200,0.0488737788465288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,65536,0.06678489181730482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,16384,0.020600888464185927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,51200,0.049044443501366504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,5120,0.012872888810104795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,51200,0.05232888791296217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,16384,0.01964000033007728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,16384,0.020279111133681405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,10240,0.017786666750907898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,32,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,12288,0.016694222887357075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,8192,0.01590488851070404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,10240,0.014840889308187695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,65536,0.06174311372968885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,7168,0.013824000126785703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,8192,0.011776000261306763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,8192,0.012804444465372296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,6144,0.012772444221708508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,32,0.0032026666320032547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,7168,0.011196444431940714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,12288,0.016760889026853774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,7168,0.011811555259757571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,6144,0.010425777898894416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,4096,0.010406222608354356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,256,0.0034968890249729156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,5120,0.009702222214804756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,4096,0.008928889201747047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,4096,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,3584,0.00849866701496972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,3584,0.008048888709810045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,5120,0.011115555961926779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,3072,0.007691555553012424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,2560,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,3584,0.008535111116038429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,2560,0.007320889168315464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,2048,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,3072,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,2048,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,1536,0.00554933316177792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,6144,0.010632889138327705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,1536,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,10240,0.014142221874660917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,1024,0.00423555572827657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,2560,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,768,0.004218666503826777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,1024,0.005971555494599872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,768,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,512,0.003579555700222651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,5120,0.009703999592198266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,512,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,1536,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,256,0.0035502223504914176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,1024,0.006607111129495833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,256,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,128,0.0032346666687064697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,256,0.004967111266321606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,2048,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,64,0.0028817777832349143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,12288,0.01590399940808614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,32,0.0032186667538351486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,64,0.005212444398138259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,65536,0.043663998444875084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,32,0.005306666509972678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,65536,0.06373511420355903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,51200,0.03559822175237868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,768,0.005614222337802251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,3072,0.008055999875068665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,51200,0.047925333182017006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,16384,0.016184889607959323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,512,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,12288,0.013742222554153867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,16384,0.01994400057527754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,12288,0.014910222755538093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,65536,0.05977866384718153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,10240,0.013785777820481194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,128,0.005974222388532426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,128,0.004969777746333016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,10240,0.01346577786737018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,7168,0.012691555751694573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,8192,0.011975111232863532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,8192,0.012150222228633033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,6144,0.011383111278216044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,7168,0.011043555206722684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,7168,0.011418666276666852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,5120,0.009725333087974125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,8192,0.012424888710180918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,5120,0.009383999639087254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,10240,0.014226666755146451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,16384,0.018055111169815063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,4096,0.008387555678685507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,6144,0.010392888552612728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,5120,0.00962311112218433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,4096,0.008539555801285638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,3072,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,4096,0.008671110702885522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,3072,0.007584888901975419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,12288,0.015928000211715698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,2560,0.006342222293217977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,2560,0.007687110867765214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,3584,0.008071111308203803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,2048,0.005588444570700328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,51200,0.05045688814587063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,1536,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,2048,0.007293333609898885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,2048,0.006676444576846228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,1024,0.004228444563017951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,1536,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,1024,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,768,0.003920888735188378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,1024,0.005664000080691443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,768,0.006329777754015393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,512,0.003528000166018804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,768,0.005583999885453119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,3584,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,256,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,512,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,512,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,128,0.0029004443850782183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,256,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,256,0.005046222358942032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,64,0.002899555489420891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,128,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,32,0.0029111111329661477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,128,0.004977777600288391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,768,65536,0.03442044390572442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,64,0.004938666605287128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,2560,0.007133333219422235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,65536,0.05824266539679634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,65536,0.062352001667022705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,6144,0.010388444695207808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,768,16384,0.014659555421935188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,51200,0.048956443866093956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,3072,0.008012444608741337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,768,12288,0.012777778009573618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,16384,0.017667555146747164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,16384,0.019322666856977675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,12288,0.014831999937693277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,12288,0.01550044450494978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,768,51200,0.027870221270455256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,10240,0.013454222016864352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,10240,0.01347733371787601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,1536,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,8192,0.01178311142656538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,8192,0.012103111379676394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,7168,0.011062222222487131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,7168,0.011048888994587792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,768,6144,0.010759111079904767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,768,10240,0.011752000285519494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,768,5120,0.009355555805895064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,32,0.005087999833954705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,6144,0.010330666270520952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,768,4096,0.007976888782448238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,6144,0.01038577738735411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,5120,0.009580444130632613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,5120,0.009196444518036312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,4096,0.00867733359336853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,51200,0.04685777756902906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,768,7168,0.012100444071822695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,768,3072,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,3584,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,3584,0.00795111142926746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,768,2560,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,3072,0.007991111112965478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,3072,0.007612444460391998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,768,2048,0.005281777845488654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,2560,0.0076657773719893555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,2560,0.0069875551594628235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,768,1536,0.004578666554556953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,2048,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,2048,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,768,1024,0.0038844446341196695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,4096,0.008343111309740279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,3584,0.007556444240940942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,768,768,0.0035786665976047516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,1024,0.006647999915811751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,1024,0.005730666634109285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,768,512,0.0032248888164758682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,768,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,768,0.005655999812814925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,768,256,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,512,0.005382222019963794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,768,8192,0.01073066641887029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,768,128,0.002967111145456632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,256,0.005970666805903117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,1536,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,768,64,0.0028560000161329904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,1536,0.006304889089531369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,128,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,768,32,0.0028844444702068963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,128,0.00498844434817632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,512,65536,0.026103999879625108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,512,51200,0.02271466619438595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,32,0.0052391112678580815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,65536,0.06103733513090345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,512,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,51200,0.04518489042917887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,51200,0.04829244481192695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,512,12288,0.013582222163677216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,16384,0.01755644381046295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,768,3584,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,512,10240,0.012522666818565793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,12288,0.014765333798196582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,12288,0.01513155632548862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,512,8192,0.011131555669837527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,64,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,65536,0.05605866511662801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,512,7168,0.010399111443095738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,8192,0.011747555600272285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,8192,0.011774222056070963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,512,16384,0.014628445108731588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,7168,0.011060444845093621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,7168,0.011069333387745751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,256,0.005200888961553574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,6144,0.010424888796276517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,5120,0.009406222237481011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,10240,0.013463999662134381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,10240,0.013774221969975365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,4096,0.008685333033402761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,4096,0.008325333396593729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,512,3584,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,3584,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,512,3072,0.006659555352396435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,512,4096,0.007852444218264686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,3072,0.008047999607192146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,512,2560,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,512,6144,0.009383999639087254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,2560,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,2560,0.007024889190991719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,512,2048,0.005250666704442766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,2048,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,2048,0.006691555596060223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,5120,0.009214222431182861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,1536,0.007019555403126611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,512,1024,0.0041680000722408295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,1024,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,3584,0.008007110820876228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,512,768,0.0036026665733920205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,1024,0.006656000183688269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,512,1536,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,3072,0.007601777712504069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,512,512,0.003583999971548716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,768,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,512,256,0.003015111097031169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,1536,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,512,0.0053119998839166425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,512,128,0.002874666617976295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,256,0.006008000009589725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,256,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,512,64,0.0028453332682450614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,16384,0.01853155593077342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,128,0.00600355573826366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,512,32,0.00258933338854048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,128,0.004933333231343163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,256,65536,0.024492444263564214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,64,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,768,0.006331555545330048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,256,51200,0.02167644434505039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,65536,0.054767999384138316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,512,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,256,16384,0.01074044406414032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,51200,0.04387199878692627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,51200,0.046542220645480685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,256,12288,0.010777778095669217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,16384,0.018253333038753934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,12288,0.014495111174053617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,12288,0.014865777558750577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,32,0.004961777892377642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,65536,0.05782133340835571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,256,8192,0.010047111246320937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,10240,0.013450666434235044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,8192,0.011640888949235281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,256,7168,0.009471110999584198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,16384,0.017284444636768766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,6144,0.010064889159467485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,256,6144,0.009000889129108852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,7168,0.011065777805116443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,7168,0.011090666883521609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,6144,0.010118222071064843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,6144,0.010040889183680216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,10240,0.013450666434235044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,5120,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,4096,0.008685333033402761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,512,5120,0.009349333743254343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,256,10240,0.009720889230569204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,256,5120,0.008719111482302347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,256,3072,0.006608888920810487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,3584,0.00832799987660514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,5120,0.009400000174840292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,3072,0.008018666671382057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,256,2560,0.005702222386995952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,3072,0.007357333269384172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,256,2048,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,2560,0.007634667058785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,2560,0.006937777830494775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,2048,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,2048,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,3584,0.007641777396202087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,1536,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,256,4096,0.007668444679843054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,1024,0.005677333308590784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,256,768,0.0038853333228164246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,1024,0.006677333265542984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,8192,0.011766222616036734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,256,512,0.003220444545149803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,256,3584,0.007288888924651676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,768,0.006342222293217977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,256,1536,0.00464088883664873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,256,256,0.0032239999208185407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,768,0.00562755556570159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,512,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,256,128,0.0028560000161329904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,512,0.0053075556125905775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,256,64,0.002512000087234709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,256,1024,0.004226666771703296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,256,32,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,128,0.004953777624501122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,128,65536,0.016191111670600045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,64,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,32,0.005033777819739448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,128,51200,0.013788444300492605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,65536,0.05181599987877739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,4096,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,128,16384,0.009704888694816166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,51200,0.04177422324816386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,51200,0.04620888829231262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,256,0.005973333285914526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,128,12288,0.012423999607563019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,16384,0.017204445269372728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,128,0.005986666513813867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,128,10240,0.011354666617181567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,12288,0.014631110760900708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,128,8192,0.010042666561073726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,10240,0.013181333740552267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,10240,0.01351377781894472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,8192,0.01147555559873581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,8192,0.011591110792424945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,7168,0.011110222174061669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,7168,0.01108000013563368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,128,6144,0.008671999805503422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,1536,0.005962666538026597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,16384,0.018225777480337355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,128,5120,0.008096888661384583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,6144,0.010033778018421596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,128,4096,0.007638221813572778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,5120,0.009348444640636444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,5120,0.00904088881280687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,128,3584,0.006871999965773688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,65536,0.05848889218436348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,3584,0.008336000144481659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,3584,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,128,3072,0.006269333263238271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,256,0.004912000149488449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,6144,0.010088889135254754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,128,2560,0.005614222337802251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,3072,0.007374222079912822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,12288,0.014793778459231058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,2560,0.007661333514584436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,2560,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,128,7168,0.009413333402739631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,128,1536,0.00444088876247406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,2048,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,2048,0.006607111129495833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,1536,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,1536,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,3072,0.008000888758235509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,1024,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,1024,0.005678222411208683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,768,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,128,2048,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,128,512,0.00319377767542998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,768,0.005698666804366642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,4096,0.0086666668454806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,4096,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,512,0.005320000151793162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,128,128,0.003071111109521654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,256,0.005247999810510212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,128,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,128,64,0.0028239999794297745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,64,0.004956444518433677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,128,256,0.0029102222373088202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,128,32,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,512,0.006028444402747684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,64,65536,0.014470222923490735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,64,51200,0.022335999541812476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,64,16384,0.009731555150614845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,32,0.00498577786816491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,65536,0.057449777921040855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,51200,0.045540445380740695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,64,10240,0.009007111191749573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,16384,0.01789066692193349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,64,8192,0.008235555556085374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,12288,0.014841778410805596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,64,7168,0.008894222478071848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,10240,0.013225778109497495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,64,6144,0.008378666308191087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,8192,0.011672889192899069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,64,5120,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,7168,0.01072622256146537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,128,1024,0.003924444317817688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,6144,0.010040000081062317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,64,12288,0.009870222045315636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,5120,0.009026666482289633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,64,3072,0.006261333409282897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,128,0.004948444250557157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,64,2560,0.005619555711746216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,3584,0.007763555480374231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,64,2048,0.005274666680230035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,3072,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,64,1536,0.004229333251714706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,2560,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,64,1024,0.003863111138343811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,2048,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,256,0.005977777971161737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,1536,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,64,512,0.0032400000426504347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,64,3584,0.0069546666410234236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,64,256,0.0029013332807355454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,4096,0.008045333127180735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,64,128,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,256,0.005000888887378904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,64,64,0.0025519999778932994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,128,0.004951111144489712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,64,32,0.0028773333049482773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,64,0.005160000175237656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,32,65536,0.016192888220151264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,1024,0.005667555663320754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,32,51200,0.024785778588718835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,128,768,0.0035502223504914176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,32,16384,0.01553422212600708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,65536,0.05662755833731758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,32,12288,0.013117333253224691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,51200,0.04535288943184746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,32,10240,0.011763555308183035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,16384,0.01796888808409373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,32,8192,0.008175111479229396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,12288,0.01482399966981676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,64,768,0.003547555456558863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,10240,0.013219555219014486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,32,6144,0.008361777497662438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,8192,0.011406222151385413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,32,5120,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,768,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,32,4096,0.007376888559924231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,6144,0.010103111465771994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,32,3584,0.006659555352396435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,5120,0.008911111288600499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,32,3072,0.006239111224810283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,4096,0.008016000191370646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,32,2560,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,3584,0.007915555602974361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,64,4096,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,32,2048,0.004959110998445087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,3072,0.007100444700982835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,32,1536,0.004259555704063839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,32,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,32,1024,0.0038773334688610504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,7168,0.011062222222487131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,32,768,0.003600888782077365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,1024,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,32,512,0.003245333416594399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,768,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,32,256,0.0031893334041039148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,32,7168,0.009026666482289633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,32,128,0.0028444443725877335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,256,0.0052728888889153796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,2560,0.006981333096822103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,2048,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,32,32,0.0025271111064487034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,64,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,32,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,512,0.005280888742870755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,257,32,64,0.0028391112056043413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,1536,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,16384,0.499616887834337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,51200,1.5298906962076824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,51200,1.1008524364895291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,128,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,16384,0.14578133159213597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,51200,0.4270266691843669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,12288,0.2571937772962782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,12288,0.1079982254240248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,10240,0.24795201089647081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,512,0.005250666704442766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,10240,0.2206266721089681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,16384,0.3500266604953342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,8192,0.21294311682383218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,12288,0.24822844399346244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,10240,0.09454666905932957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,16384,0.3217724429236518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,8192,0.17641599973042807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,8192,0.07628799809349908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,10240,0.20138577620188394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,7168,0.1571093267864651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,7168,0.06791200240453084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,6144,0.15345066123538548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,8192,0.16310844156477186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,6144,0.13570666313171387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,6144,0.05787377887301975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,7168,0.1436986658308241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,7168,0.2118551068835788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,6144,0.1274577776590983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,5120,0.1135680013232761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,4096,0.11508444945017497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,12288,0.37568267186482746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,4096,0.094760000705719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,5120,0.10454399718178643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,5120,0.13866310649447972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,3584,0.08220444122950236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,3584,0.03733511103524102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,51200,1.00418578253852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,3584,0.07545777824189928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,3072,0.07030400302675036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,2560,0.07547377877765231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,4096,0.04284711016549004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,3072,0.06957866748174031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,2560,0.03010666701528761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,2560,0.05884266561932034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,3072,0.08591644631491767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,5120,0.04951733350753784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,3072,0.03353510962592231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,2048,0.05139644609557258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,1536,0.04405511087841458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,3584,0.09440444575415717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,4096,0.08615466621186997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,1024,0.03331289026472304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,1536,0.04036710990799798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,1536,0.020676443974177044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,1536,0.03973066806793213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,2048,0.05883644686804878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,768,0.026388444834285315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,1024,0.03132799930042691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,1024,0.0320995549360911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,768,0.014134221606784396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,512,0.019856888386938307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,768,0.025965332984924316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,768,0.027436445156733196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,2048,0.0483439995182885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,512,0.01310400002532535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,512,0.02129688858985901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,512,0.02374222212367588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,128,0.011752888560295105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,256,0.012183999849690331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,256,0.018939556346999276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,256,0.019496000475353666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,64,0.009508444203270806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,1024,0.01573066744539473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,128,0.01184088902340995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,32,0.009860444400045607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,128,0.017383999294704862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,64,0.01791200041770935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,2048,0.025983999172846477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,256,0.013154666456911298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,2560,0.06024622254901462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,128,0.01812977757718828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,65536,1.3520328733656142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,65536,1.0959350797865126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,51200,1.0473768446180556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,32,0.017906667457686532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,51200,0.8431653446621366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,16384,0.34206755956013996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,16384,0.27761244773864746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,12288,0.2521786689758301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,51200,0.33951200379265684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,16384,0.11340622107187907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,12288,0.20132622453901503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,12288,0.08586311340332031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,16384,0.2591804398430718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,10240,0.17167466216617158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,12288,0.19732888539632162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,8192,0.17876177363925508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,8192,0.14052622848086888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,10240,0.16433333026038274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,8192,0.059862222936418324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,7168,0.15476977825164795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,8192,0.13794755935668945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,7168,0.12420711252424453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,65536,0.4260888894399007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,7168,0.054382221566306226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,6144,0.11008177863226996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,10240,0.1751555601755778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,7168,0.11754577689700657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,6144,0.046655111842685275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,6144,0.10525421963797675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,6144,0.10509955883026123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,5120,0.08943644497129653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,65536,1.0147155125935872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,4096,0.09089599715338813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,5120,0.0398248897658454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,5120,0.08636533551745945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,3584,0.07736444473266602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,4096,0.07317599985334608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,5120,0.09370755487018162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,3584,0.030626667870415583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,3584,0.06240799691942003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,3072,0.05557866891225179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,51200,0.854881763458252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,3072,0.05579644441604614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,2560,0.05101955599255032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,3072,0.02772088845570882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,2560,0.048455112510257296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,2048,0.04560533497068617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,3072,0.058876441584693066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,2048,0.041322665082083807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,10240,0.07243200143178304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,1536,0.03477688961558872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,2048,0.020793777373101976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,4096,0.03456622362136841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,2560,0.024706666668256123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,1536,0.03312355611059401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,1024,0.0251511103577084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,4096,0.07295999924341838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,1536,0.03467288944456313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,768,0.019644444187482197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,1024,0.014125333891974555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,1024,0.02560088866286808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,1024,0.028359111812379625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,768,0.013044444223244986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,2048,0.04169155491722954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,768,0.024383111132515803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,512,0.01759111053413815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,512,0.011761777930789523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,1536,0.017009778155220878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,2560,0.048335111803478666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,3584,0.06330844428804186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,128,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,256,0.010501333408885531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,256,0.016180444094869826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,256,0.01664888858795166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,64,0.009190222455395592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,128,0.01071911139620675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,512,0.014916444818178812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,768,0.0207342223988639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,128,0.01552888916598426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,64,0.014904888139830695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,32,0.015074667003419666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,256,0.011679111255539788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,65536,0.3917555544111464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,512,0.02100533412562476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,65536,0.367556439505683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,51200,0.3061777750651042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,65536,0.14660444524553087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,16384,0.10527644554773967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,128,0.015152888165579902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,32,0.008984000318580205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,12288,0.07991821898354425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,16384,0.04217155443297493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,51200,0.11855199601915146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,16384,0.0882906648847792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,12288,0.07084621985753377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,12288,0.03280266788270738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,12288,0.0683635539478726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,51200,0.2745199998219808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,10240,0.02846133377816942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,8192,0.05863822168774075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,65536,0.3548915651109483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,51200,0.28941867086622447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,16384,0.09308266639709473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,7168,0.051615112357669406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,10240,0.06832089026769002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,8192,0.02421955598725213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,7168,0.043895999590555825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,6144,0.044271111488342285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,10240,0.059643553362952344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,7168,0.02165777815712823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,7168,0.044938666952980884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,5120,0.03684533304638333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,6144,0.03854044609599643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,6144,0.04082666503058539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,5120,0.03304800060060289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,4096,0.031126221021016438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,5120,0.017108443710539076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,8192,0.050873776276906334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,3584,0.028284443749321833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,4096,0.027931554449929133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,10240,0.05894044372770521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,6144,0.019408888287014432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,3584,0.024937777055634394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,3072,0.024864888853496973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,3584,0.02755644420782725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,3072,0.01314222233162986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,5120,0.035522666242387556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,3072,0.022203556365436975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,3072,0.025383111503389146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,4096,0.015524443652894763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,2560,0.019369777705934312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,2048,0.017470222380426195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,2560,0.012221333053376941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,8192,0.05002311203214857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,3584,0.0140328887436125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,1536,0.014153778553009033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,2048,0.016970665918456185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,2048,0.01109866632355584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,2048,0.020192889703644645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,1024,0.011384888655609555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,1536,0.009383111364311641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,1536,0.014386667145623101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,2560,0.020764445265134174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,1024,0.007985778152942657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,1024,0.014927999840842353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,768,0.010108444425794814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,4096,0.03065511253145006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,512,0.008368000388145447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,768,0.013655111193656921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,2560,0.023043556345833674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,256,0.005595555735958948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,512,0.01146755533085929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,768,0.009701333112186855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,1024,0.011902222202883827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,128,0.004934222333961063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,768,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,128,0.008237332933478886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,512,0.00741511086622874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,64,0.0048071112897661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,128,0.010031999813185798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,32,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,1536,0.017679111825095285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,64,0.008205333517657386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,256,0.006980444822046492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,32,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,65536,0.30782222747802734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,51200,0.2309866746266683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,65536,0.29391998714870876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,65536,0.1289093361960517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,256,0.00814577771557702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,51200,0.10277333524492051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,16384,0.08940000004238552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,256,0.01037688884470198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,16384,0.03871022330390082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,12288,0.06685066885418363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,65536,0.3197253280215793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,16384,0.07777422004275851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,512,0.008735111190213097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,51200,0.2470222314198812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,10240,0.05338489015897115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,16384,0.08621511194441055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,12288,0.05939200189378527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,8192,0.04624888963169522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,12288,0.06694933440950182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,51200,0.22913244035508898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,10240,0.05068177647060818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,7168,0.039988444911109075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,8192,0.02202044427394867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,8192,0.04234310984611511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,8192,0.04833066794607374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,7168,0.01993600030740102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,7168,0.03752622339460585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,128,0.007009777757856581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,7168,0.043382220798068576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,5120,0.02975822157329983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,12288,0.030696001317765977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,6144,0.03894844320085313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,5120,0.028202666176689997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,4096,0.024930665890375774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,5120,0.034161776304244995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,6144,0.033377776543299355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,4096,0.023815110325813293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,4096,0.01480888823668162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,3584,0.021908443835046556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,4096,0.029312888781229656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,10240,0.025797333982255723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,3584,0.021456888980335657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,5120,0.016389333539538913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,3072,0.01919555498494042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,3584,0.0266977780395084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,3072,0.012717333104875354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,2560,0.01646488904953003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,6144,0.01835644410716163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,2560,0.017095110482639737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,10240,0.05839733282725016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,2560,0.02228533393806881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,2048,0.014064888159434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,3072,0.019336889187494915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,3072,0.024757333927684363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,2048,0.019934222102165222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,1536,0.010934222075674268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,2048,0.010240000155236986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,6144,0.033032000064849854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,1024,0.009697777529557547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,1536,0.00870666652917862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,1536,0.01276266657643848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,768,0.008522666990756989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,1024,0.007835555407736037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,3584,0.013696889082590738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,1024,0.014056889547242058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,512,0.006726222319735422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,768,0.009015999734401703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,768,0.007753777835104201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,2048,0.015027556154463025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,256,0.005234666582610872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,512,0.007420444654093848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,512,0.010768888725174798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,512,0.0081395556529363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,256,0.007442666424645319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,256,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,1536,0.017260443833139207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,1024,0.010372444159454769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,256,0.01016533292002148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,64,0.004562666846646203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,128,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,128,0.009712888962692684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,64,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,32,0.007450666692521837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,65536,0.2293093204498291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,128,0.004648000001907349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,65536,0.2511315610673693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,2560,0.011694221860832639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,51200,0.22327287991841635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,128,0.007363555332024892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,16384,0.07030755281448364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,768,0.012810666528013019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,51200,0.19596266746520996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,51200,0.10022844208611383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,12288,0.05367111166318258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,65536,0.3144133355882433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,16384,0.03773777683575948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,16384,0.08442311154471503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,10240,0.043509334325790405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,65536,0.12486399544609918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,12288,0.02903555499182807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,12288,0.051382223765055336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,8192,0.03681688838534885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,10240,0.02494311167134179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,10240,0.05767644113964505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,16384,0.0662577748298645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,8192,0.02144444485505422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,8192,0.03699288765589396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,51200,0.24497334162394205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,7168,0.03264177838961283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,32,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,7168,0.01955200069480472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,7168,0.04246755441029867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,10240,0.04395111070738899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,5120,0.02369155486424764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,6144,0.028456889920764502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,6144,0.03801244497299194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,7168,0.03200622068511115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,5120,0.024431111084090337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,5120,0.015826667348543804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,5120,0.03359555535846286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,8192,0.047576887740029224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,3584,0.017423111531469557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,4096,0.014427555931939019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,4096,0.020743111769358318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,4096,0.02846755584081014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,3072,0.015484443969196744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,3584,0.01344622257683012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,3584,0.018834667073355783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,6144,0.017699556218253244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,3584,0.026309332913822595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,2560,0.013487999637921652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,3072,0.0123857781291008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,3072,0.01711466742886437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,2048,0.01166488892502255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,2560,0.015384000208642749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,2560,0.011001778145631155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,2048,0.013456888496875763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,2560,0.02127200033929613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,1536,0.010022222167915767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,2048,0.00908088849650489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,2048,0.019147555033365887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,1536,0.0166320006052653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,3072,0.024108444650967915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,12288,0.06627288791868421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,1024,0.009473777479595607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,1024,0.007701333198282454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,768,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,1024,0.013456000222100152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,4096,0.020310221446885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,768,0.012080888781282636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,512,0.005376889059940974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,6144,0.02755022214518653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,512,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,256,0.004564444637960858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,512,0.010715555813577441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,256,0.006980444822046492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,1536,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,128,0.004106666478845808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,256,0.006949333681000604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,128,0.006949333681000604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,512,0.007034666836261749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,128,0.009707555174827576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,128,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,32,0.004276444514592488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,1024,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,768,0.008350222474998897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,256,0.009659555223253038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,768,0.007341333561473423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,32,0.007075555622577667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,51200,0.1870266728931003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,65536,0.20400355921851265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,65536,0.0799964467684428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,64,0.003903999924659729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,51200,0.0637342201338874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,1536,0.011617778076065911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,65536,0.1741102271609836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,64,0.006955555743641323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,12288,0.0430026650428772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,16384,0.054995556672414146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,16384,0.0243857784403695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,16384,0.04849422309133741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,10240,0.03791200121243795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,12288,0.042466666963365346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,12288,0.039557334449556135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,51200,0.1594453387790256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,8192,0.032990223831600614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,10240,0.0341635545094808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,65536,0.20639467239379883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,51200,0.13720977306365967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,8192,0.030568887790044148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,7168,0.0281315545241038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,8192,0.029245333539115057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,8192,0.01513777838812934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,12288,0.019490665859646268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,6144,0.02475555572244856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,7168,0.014109333356221518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,7168,0.02680355476008521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,6144,0.012991999586423239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,10240,0.03664355476697286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,6144,0.024023110667864483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,5120,0.011943110989199745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,4096,0.017934223016103108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,5120,0.021998221675554912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,4096,0.017850667238235474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,7168,0.02702666653527154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,4096,0.019278221660190158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,4096,0.011002666420406766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,5120,0.020784000555674236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,16384,0.05936355723275078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,3072,0.014022222823566861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,3584,0.010069333016872406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,5120,0.020844444632530212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,3584,0.018070222602950204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,2560,0.01237511138121287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,3072,0.00903555585278405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,3072,0.016920000314712524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,2560,0.013447110851605734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,2560,0.008719111482302347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,2048,0.01071733319097095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,2560,0.01566133399804433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,6144,0.02403644389576382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,2048,0.007985778152942657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,2048,0.012048888537618848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,2048,0.013870221873124441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,3072,0.015051555302408008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,1024,0.006950221955776215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,1536,0.012251555919647217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,1024,0.008361777497662438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,1024,0.01073155552148819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,10240,0.017238222890430026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,768,0.005712888720962737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,1536,0.008660444782839881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,3584,0.01628444426589542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,512,0.0046924443708525766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,768,0.006650666809744305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,768,0.007548444800906711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,1536,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,768,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,256,0.004278222305907143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,512,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,512,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,512,0.00906399968597624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,256,0.006256000035338932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,256,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,256,0.008348444269763099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,1024,0.006758222149478064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,64,0.003607999947335985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,128,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,32,0.0038764443662431505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,128,0.008464888566070134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,1536,0.010088000032636855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,65536,0.18976888391706678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,128,0.003914666672547658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,65536,0.2012444469663832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,51200,0.14120622475941977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,128,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,64,0.006611555400821898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,51200,0.1565342214372423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,16384,0.0481333335240682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,51200,0.059980445437961154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,3584,0.015480889214409722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,51200,0.1268035570780436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,12288,0.040616889794667564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,16384,0.05359289050102234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,16384,0.023471110396915015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,16384,0.04716888732380337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,65536,0.0748035576608446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,10240,0.03511910968356662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,12288,0.04207466708289253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,8192,0.028410666518741187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,65536,0.16234578026665583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,10240,0.035903112755881414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,10240,0.016910221841600206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,12288,0.03827022181616889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,7168,0.025615110993385315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,8192,0.014870221416155497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,8192,0.029073778125974867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,7168,0.013657777673668332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,6144,0.021347555849287245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,7168,0.026424888107511733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,12288,0.018811555372344125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,7168,0.02661955522166358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,5120,0.018612444400787354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,6144,0.012792000340090858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,6144,0.024297777149412367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,4096,0.015205333630243937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,32,0.006640000061856375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,10240,0.0337413317627377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,5120,0.011718221836619906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,8192,0.02988977895842658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,5120,0.021584888299306233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,5120,0.020444444484180875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,4096,0.01757777730623881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,3072,0.012702222499582501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,4096,0.01870577699608273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,3584,0.016179554992251925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,3584,0.01759288873937395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,2560,0.011325333681371478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,3072,0.014811555544535318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,3072,0.016520889268981088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,2560,0.008310221963458592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,2048,0.009711110757456886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,3584,0.014160888062583076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,4096,0.010310221877362993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,3584,0.009573332965373993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,2048,0.007637333538797166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,1536,0.00851733320289188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,2048,0.013901333014170328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,1536,0.009383999639087254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,1024,0.006657777975002925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,1536,0.007148444652557373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,2560,0.01309511148267322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,2560,0.015501333607567681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,768,0.005984888722499211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,1024,0.006674666785531574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,1024,0.008031110796663497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,1024,0.010051555931568146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,768,0.006671999891599019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,512,0.005095110999213325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,768,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,768,0.009691555466916826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,6144,0.023476445012622412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,512,0.006325333482689328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,512,0.0069573331210348345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,512,0.00868088917599784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,128,0.004247111164861255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,256,0.006320888797442119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,256,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,1536,0.011807999677128263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,64,0.0039813332259655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,128,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,128,0.008720888859695857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,2048,0.011761777930789523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,65536,0.1639208926094903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,256,0.004549333204825719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,32,0.006698666761318843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,65536,0.1620186699761285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,128,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,51200,0.1457564433415731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,32,0.003957333250178231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,65536,0.1543697781032986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,64,0.0063475556671619415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,3072,0.008660444782839881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,16384,0.049192888869179614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,51200,0.057654221852620445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,16384,0.021925333473417494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,16384,0.04433600107828776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,12288,0.03814222084151374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,16384,0.046888000435299344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,12288,0.03449155555831061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,65536,0.07089955276913114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,10240,0.03126044405831231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,12288,0.018231999542978074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,8192,0.026133333643277485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,10240,0.016168889072206285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,51200,0.12495911121368408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,10240,0.029547555579079524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,7168,0.02312622302108341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,8192,0.014483556151390076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,51200,0.12309066454569499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,7168,0.022287999590237934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,7168,0.013399110900031196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,7168,0.026075555218590632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,12288,0.03789510991838243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,6144,0.019811555743217468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,6144,0.012440000143316058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,5120,0.017027555240525138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,6144,0.023582221733199224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,10240,0.033560888634787664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,8192,0.02464888824356927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,5120,0.017362667454613578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,256,0.008727110922336578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,5120,0.021352888809310064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,3584,0.013089777694808112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,4096,0.009728888670603434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,4096,0.01516711049609714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,4096,0.01872977779971229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,3072,0.011760888828171624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,3584,0.014024888475735983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,3584,0.017619555195172627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,3072,0.012795555922720166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,2560,0.010747555229398938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,4096,0.014663111832406787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,8192,0.028583112690183852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,2560,0.011771555576059552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,2048,0.009307555854320526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,3584,0.009025777379671732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,2048,0.010065777434243096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,3072,0.008963555925422246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,2048,0.00740622232357661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,5120,0.011075555450386472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,1536,0.008349333372380998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,1536,0.011765333513418833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,1024,0.006037333359320958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,1536,0.0070088886552386815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,1024,0.007320889168315464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,3072,0.016255110502243042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,1024,0.006690666493442323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,768,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,2560,0.007976888782448238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,2560,0.015437333120240105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,512,0.004613333278232151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,768,0.006704000135262807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,768,0.009445333646403419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,512,0.006311999840868845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,256,0.00424799985355801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,2048,0.013110222087966071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,512,0.0063768890168931745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,512,0.009496000077989366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,256,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,256,0.006225777582989798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,6144,0.019928889142142404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,128,0.006132444573773279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,64,0.0037048889531029593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,1024,0.010083555347389644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,64,0.005974222388532426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,32,0.0038071109188927543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,65536,0.1461120049158732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,128,0.0038888889054457345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,32,0.006235555642180973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,128,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,65536,0.06857955455780029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,128,0.008041777544551426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,51200,0.12626400258806017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,1536,0.007621333003044128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,65536,0.15370133188035753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,51200,0.12253777186075847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,256,0.008350222474998897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,51200,0.12267643875545925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,16384,0.04638933473163181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,12288,0.03282311227586534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,16384,0.02124711043304867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,65536,0.15550488895840114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,10240,0.026761778526835974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,12288,0.01798311041461097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,12288,0.03303466571701898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,12288,0.03711555401484171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,51200,0.055845333470238574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,16384,0.04187733266088697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,10240,0.028513779242833454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,10240,0.015824000040690105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,10240,0.0327413347032335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,7168,0.019820445113711886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,8192,0.024101333485709295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,16384,0.04266666703754001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,7168,0.021547555923461914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,7168,0.012863111164834766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,6144,0.017436444759368896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,7168,0.026164443956481084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,6144,0.01203733351495531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,6144,0.019537778364287484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,6144,0.023200000325838726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,768,0.006913777854707506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,5120,0.017027555240525138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,4096,0.013144888811641268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,5120,0.020806221498383414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,4096,0.014749333262443542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,3584,0.011995555626021491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,8192,0.022843556271659002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,3584,0.013549333645237817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,8192,0.013903111219406128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,3072,0.010896888871987661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,3584,0.008674666285514832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,4096,0.009022222624884712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,3072,0.012759110993809171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,5120,0.01516888870133294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,3072,0.016357334123717416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,2560,0.01075111081202825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,3584,0.017494221528371174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,8192,0.028074665202034846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,2560,0.007887110941939885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,2048,0.009729777773221334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,4096,0.018237334158685472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,2048,0.013152889079517789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,2560,0.009759110709031424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,2048,0.007409777906205919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,1536,0.007010666860474481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,1024,0.005640000104904175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,1536,0.011405333048767514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,5120,0.010746666954623329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,1024,0.006610666712125142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,1024,0.007297777467303806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,1024,0.010062221851613786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,1536,0.007252444823582967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,768,0.006703111032644908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,768,0.0063733334342638654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,768,0.009011555876996782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,1536,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,512,0.0063528890411059065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,256,0.003919110943873723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,512,0.008747555315494537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,768,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,256,0.006062222023804982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,128,0.003547555456558863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,3072,0.008053333395057255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,512,0.004256000121434529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,128,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,64,0.003528000166018804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,2560,0.015077332655588785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,32,0.003646222253640493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,64,0.006025777922736273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,32,0.006034666465388403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,65536,0.12343022558424209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,256,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,128,0.00598933340774642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,65536,0.06645333104663424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,65536,0.11811378267076279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,128,0.008332444561852349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,65536,0.14617244402567545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,16384,0.03512977891498142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,51200,0.05298755566279093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,256,0.005998222364319696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,16384,0.0325617790222168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,12288,0.026937777797381084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,51200,0.11618578433990479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,16384,0.04360266526540121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,12288,0.01715377800994449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,12288,0.025702221526039973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,51200,0.09448888566758896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,12288,0.03527999917666117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,10240,0.015325332681337992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,10240,0.02238222294383579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,8192,0.019131556153297424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,10240,0.030702223380406696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,51200,0.09271733628378974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,7168,0.016685333516862657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,8192,0.013487999637921652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,8192,0.019312888383865356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,7168,0.012144889268610211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,8192,0.02665155629316966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,10240,0.021965333157115515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,7168,0.017675555414623685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,6144,0.011208888557222156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,6144,0.01591466698381636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,5120,0.012721777790122561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,16384,0.020632889535692003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,6144,0.014628445108731588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,5120,0.014124444789356656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,2048,0.008409777449236976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,5120,0.020104888412687514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,5120,0.009720889230569204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,4096,0.009033777647548253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,4096,0.012108444339699216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,3584,0.010300444232092964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,4096,0.017431111799346078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,3584,0.008449777960777283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,3584,0.011120888921949597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,3584,0.01625066664483812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,512,0.006309333360857434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,3072,0.015263110399246216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,3072,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,2560,0.00865955568022198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,6144,0.0217866665787167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,2560,0.007777777810891469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,2560,0.013812444276279874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,3072,0.009738666315873465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,2048,0.008171555896600088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,2048,0.011884444289737277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,2048,0.007334222396214803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,1536,0.007369777394665613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,1536,0.010715555813577441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,2560,0.009098666409651438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,1536,0.0070088886552386815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,2048,0.007675555845101674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,1024,0.006614222294754452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,768,0.005019555489222209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,1024,0.009760888914267222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,1536,0.006979555719428592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,768,0.00638133328821924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,512,0.004507555729813046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,768,0.009354666703277165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,1024,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,512,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,256,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,1024,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,7168,0.024363555841975745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,256,0.005632888939645555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,128,0.003549333247873518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,768,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,256,0.008013333711359236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,128,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,64,0.0035217776894569397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,128,0.007976888782448238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,32,0.003552000141806073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,512,0.008716444174448649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,64,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,4096,0.01146933353609509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,32,0.00563822231358952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,256,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,65536,0.11724710464477539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,65536,0.06486755609512329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,128,0.005592888842026393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,65536,0.14622044563293457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,16384,0.035475555393430926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,51200,0.0919004413816664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,51200,0.052060445149739586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,51200,0.11563821633656819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,65536,0.10807910892698501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,12288,0.02929688824547662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,3072,0.010294222169452244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,16384,0.0427706672085656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,12288,0.01686311099264357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,51200,0.08695199754503037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,10240,0.024298666252030268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,12288,0.025455110602908667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,8192,0.02070755594306522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,10240,0.0151582227812873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,512,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,8192,0.01901333365175459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,16384,0.020258666740523446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,7168,0.018258665998776753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,8192,0.025748444928063288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,7168,0.012033777932325998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,7168,0.01737866633468204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,7168,0.02426311042573717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,12288,0.03462310963206821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,10240,0.022340445054901972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,5120,0.014124444789356656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,6144,0.010759111079904767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,10240,0.030446221431096394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,6144,0.021684444612926904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,4096,0.010822222464614444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,5120,0.009576888548003303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,16384,0.03188711073663499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,5120,0.019681778219011094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,3584,0.009319999979601966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,4096,0.012135110795497894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,6144,0.01605866683853997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,4096,0.017581333716710407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,3072,0.009020444419648912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,3584,0.008432889150248634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,3584,0.016016888949606154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,3072,0.009908444351620143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,2560,0.008415111237102086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,3072,0.014485332700941296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,5120,0.013780444032616086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,8192,0.013421333498424955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,4096,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,2048,0.0069520001610120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,2560,0.013648000028398303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,3584,0.011125333607196808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,2048,0.008288000192907121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,1536,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,2048,0.007337777978844113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,3072,0.008042666647169325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,1536,0.007035555938879649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,1536,0.007324444750944774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,1536,0.010747555229398938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,2560,0.007678222325113084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,1024,0.006719111154476802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,1024,0.009707555174827576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,768,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,6144,0.015840889679061044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,2048,0.011757333245542316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,768,0.006351111249791251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,768,0.006299555715587404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,768,0.00869777798652649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,1024,0.005285333428117964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,512,0.0058853332367208265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,2560,0.00904444439543618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,512,0.00869688888390859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,256,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,1024,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,256,0.00797333319981893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,128,0.005410666680998272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,64,0.0035377778112888336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,128,0.008012444608741337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,32,0.0035502223504914176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,256,0.003875555677546395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,64,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,65536,0.09840622213151719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,32,0.005590222362014983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,128,0.003605333467324575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,65536,0.06396710872650146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,128,0.005977777971161737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,51200,0.08709155850940281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,512,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,51200,0.05121866530842251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,51200,0.11547199885050456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,256,0.005608888963858287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,16384,0.03107822272512648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,12288,0.02253599961598714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,16384,0.04318844609790378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,512,0.004218666503826777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,12288,0.016754666964213055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,10240,0.01891911029815674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,12288,0.03458933366669549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,16384,0.031395554542541504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,10240,0.014861333701345654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,8192,0.015690666106012132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,65536,0.11033955547544692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,16384,0.01994222237004174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,65536,0.14556088712480333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,8192,0.018671999375025432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,8192,0.012504888905419244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,8192,0.025792888469166223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,7168,0.011800888511869641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,7168,0.017062221964200337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,6144,0.012804444465372296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,7168,0.02390044430891673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,10240,0.021635555558734473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,6144,0.010388444695207808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,6144,0.01534400052494473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,51200,0.07556621895896064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,6144,0.021407999926143225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,4096,0.010398222340477837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,5120,0.009360888765917884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,5120,0.01368977791733212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,5120,0.01960266629854838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,12288,0.024822221861945257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,3584,0.00906044410334693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,4096,0.011195555329322815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,4096,0.016549333930015564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,3584,0.008344000412358178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,3584,0.010454222559928894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,3072,0.009723555710580613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,3584,0.015533333023389181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,2560,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,3072,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,7168,0.0141004439857271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,4096,0.00868000007337994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,3072,0.014349333114094205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,2560,0.012999999854299756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,2560,0.009022222624884712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,5120,0.011672000090281168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,2048,0.007439110842016008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,2048,0.011759999725553723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,1536,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,2560,0.007710222568776872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,2048,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,1536,0.006984000404675801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,3072,0.008292444050312042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,1024,0.006611555400821898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,10240,0.030369778474171955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,1024,0.009308444129096137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,2048,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,768,0.006183111005359226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,768,0.006322666588756774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,1536,0.010440000229411654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,768,0.009384888741705153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,1024,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,768,0.004218666503826777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,1536,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,512,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,256,0.005616888817813661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,256,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,512,0.0041519999504089355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,128,0.003552000141806073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,512,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,128,0.0053048887186580235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,64,0.0032666667054096856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,128,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,32,0.0032035555276605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,64,0.00564444437623024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,65536,0.07963644133673774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,1024,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,256,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,51200,0.06377510892020331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,65536,0.06282666656706068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,65536,0.1082728902498881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,65536,0.14520711368984646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,16384,0.02884977724817064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,51200,0.115466660923428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,512,0.008759111166000366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,16384,0.0306275553173489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,12288,0.022675555613305833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,16384,0.04308977723121643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,12288,0.024361777636739943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,128,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,10240,0.020448888341585796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,10240,0.021515554851955835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,16384,0.019705777366956074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,10240,0.014473777678277759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,10240,0.029892444610595703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,256,0.0035982223020659555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,51200,0.05008000135421753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,8192,0.018444443742434185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,7168,0.013797333670987023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,8192,0.025632888078689575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,32,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,7168,0.011065777805116443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,6144,0.0124195557501581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,7168,0.023573334018389385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,8192,0.015352000792821249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,6144,0.010379555324713389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,51200,0.08535999721950954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,6144,0.01501511037349701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,6144,0.021313778228229944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,8192,0.012422222230169507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,4096,0.009373333719041612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,5120,0.012615111139085559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,5120,0.01924711134698656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,7168,0.016542222764756944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,3584,0.008991999758614434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,4096,0.016234666109085083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,3584,0.010145777629481422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,5120,0.011033777561452655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,3072,0.008456889126035903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,5120,0.009475555684831407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,3072,0.009457777771684859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,4096,0.010784000158309937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,3072,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,12288,0.034430222378836736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,2560,0.00870400004916721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,3584,0.015068444940778943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,2560,0.007715555528799693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,2048,0.0064399999876817065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,4096,0.00889955543809467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,3072,0.014151111245155334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,1536,0.005567111074924469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,2048,0.007328888608349695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,2048,0.01146044416560067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,12288,0.0163093325164583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,1024,0.004781333522664176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,1536,0.007103111180994246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,1536,0.006989333364698622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,1536,0.010423111418883005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,768,0.004561777744028303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,1024,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,1024,0.006340444501903322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,2560,0.01310755560795466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,512,0.003940444439649582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,2048,0.008040000167157915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,768,0.008879111044936711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,512,0.0057964444988303715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,256,0.0035742223262786865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,512,0.006055111272467508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,512,0.008750222623348236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,256,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,128,0.003343111111058129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,1024,0.009369778136412302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,256,0.006037333359320958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,768,0.0063164445261160536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,64,0.0032337777730491427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,128,0.005290666802061929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,128,0.007693332930405934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,32,0.0031617778456873367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,64,0.005574222240183089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,3584,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,65536,0.06344533628887601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,32,0.005557333429654439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,65536,0.06892266538408068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,65536,0.06143110990524292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,51200,0.053435554107030235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,65536,0.14498399363623724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,256,0.008052444292439355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,128,0.005919999960396025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,16384,0.022442667020691767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,51200,0.04821866750717163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,51200,0.11519289016723633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,16384,0.01943466729587979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,12288,0.020615999897321064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,16384,0.02109066645304362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,16384,0.04276266694068909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,12288,0.015767999821239047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,12288,0.03423733181423611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,10240,0.015199111567603217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,8192,0.014448889427714877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,10240,0.014127110441525778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,8192,0.01311644415060679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,12288,0.017056889004177518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,8192,0.011690666278203329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,7168,0.013070222404268054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,768,0.006059555543793573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,7168,0.010888000329335531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,7168,0.012075555821259817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,6144,0.011847111086050669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,7168,0.0236791107389662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,6144,0.010420444111029306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,6144,0.010847999817795224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,5120,0.010231110784742568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,6144,0.021306667062971327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,8192,0.025684444440735713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,5120,0.009429333110650381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,4096,0.009183111290136972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,5120,0.01891377733813392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,4096,0.008781332936551835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,4096,0.008374222450786168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,51200,0.05419022176000807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,4096,0.016180444094869826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,3584,0.008319999608728621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,3584,0.007985778152942657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,3072,0.007634667058785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,3584,0.01516711049609714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,5120,0.009352888498041365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,2560,0.007627555893527136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,3072,0.013836444252067141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,3584,0.008365333080291748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,2560,0.007017778025733099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,10240,0.02960444490114848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,2048,0.005626666463083691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,2560,0.01311911145846049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,2560,0.007689778175618913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,1536,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,2048,0.007296000089910295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,2048,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,2048,0.011783999701340994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,1024,0.00425511101881663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,10240,0.016972444123691983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,2560,0.007285333342022366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,1536,0.010739555789364709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,768,0.00407288885778851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,1024,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,1024,0.0058577776783042485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,1024,0.009346666435400644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,512,0.003544888976547453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,768,0.006308444258239534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,768,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,768,0.00904799997806549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,256,0.003547555456558863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,512,0.005557333429654439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,512,0.008664888640244802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,3072,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,256,0.005245333330498801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,256,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,256,0.007648888561460707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,128,0.004953777624501122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,64,0.0032382222513357797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,512,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,32,0.003257777748836411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,64,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,1536,0.006989333364698622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,65536,0.05063644382688734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,128,0.003265777809752358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,32,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,51200,0.04193333453602261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,65536,0.06662133004930285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,128,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,65536,0.14428888426886663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,128,0.008021333151393467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,16384,0.018945778409639995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,51200,0.05252177847756279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,51200,0.11497155825297038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,16384,0.0203804439968533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,16384,0.018961777289708454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,1536,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,12288,0.01686755485004849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,65536,0.05986399783028496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,12288,0.033747557136747576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,51200,0.047298666503694325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,10240,0.014840000205569796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,12288,0.015841777125994366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,3072,0.008007110820876228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,8192,0.014124444789356656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,16384,0.04231111208597819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,10240,0.029672000143263075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,7168,0.01275644451379776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,8192,0.01145155562294854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,8192,0.012101333174440594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,8192,0.02533866630660163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,10240,0.014525334040323893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,6144,0.01166133334239324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,7168,0.011872889267073737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,7168,0.022980444961124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,5120,0.010208889014191097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,6144,0.010155555274751451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,6144,0.010596444209416708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,10240,0.01310488912794325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,5120,0.009526222116417354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,4096,0.008672888908121321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,5120,0.00941422250535753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,4096,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,12288,0.015420445137553744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,4096,0.016010666886965435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,3584,0.008319999608728621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,6144,0.0199608885579639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,3072,0.007634667058785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,3584,0.008000888758235509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,4096,0.008770667016506195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,2560,0.006603555546866522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,3072,0.013830222189426422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,3072,0.007975999679830339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,3584,0.008059555457697975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,2560,0.007640000018808577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,2048,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,2560,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,2560,0.0127688886390792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,1536,0.0052426668504873914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,2048,0.007294221884674496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,2048,0.011428444749779172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,1536,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,7168,0.010874666273593903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,1536,0.01070577816830741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,1024,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,3584,0.01484444406297472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,1024,0.009348444640636444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,1024,0.006640000061856375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,2048,0.006637333167923822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,768,0.005623111294375525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,512,0.003867555409669876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,768,0.008681777450773451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,5120,0.01792711185084449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,512,0.0063866666621632045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,512,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,256,0.0035617777870761026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,768,0.004212444441186057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,256,0.00592533333433999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,256,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,128,0.0031822222388452957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,256,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,768,0.006329777754015393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,64,0.0032035555276605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,128,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,1536,0.006991110742092133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,32,0.0031928889867332247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,128,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,65536,0.037913779417673744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,64,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,512,0.0053324442770746015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,3072,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,51200,0.031692445278167725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,65536,0.06376533375846015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,65536,0.14479377534654406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,16384,0.01593244406912062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,1024,0.004239999999602635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,51200,0.11480266518063015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,16384,0.01995733380317688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,12288,0.01367555558681488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,65536,0.05791999896367391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,51200,0.0466231107711792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,12288,0.015895111693276297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,10240,0.012694222231705984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,128,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,10240,0.014135110709402295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,32,0.005294222384691238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,10240,0.028885334730148315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,10240,0.012885332935386233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,51200,0.04999377661281162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,8192,0.011815999945004782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,7168,0.012411555482281579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,12288,0.03360088997417026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,7168,0.011066666907734342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,7168,0.02239111065864563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,16384,0.016900445024172466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,6144,0.01109333336353302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,7168,0.01073155552148819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,6144,0.01038044442733129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,6144,0.01995022263791826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,6144,0.010055111514197456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,5120,0.009364444348547194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,16384,0.042064001162846885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,8192,0.013800000150998434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,4096,0.008404444489214156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,5120,0.009370666411187913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,5120,0.017941332525677152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,4096,0.008651555412345463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,4096,0.008356444537639618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,4096,0.0155102229780621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,3584,0.008030222521887885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,3584,0.014803555276658801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,3072,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,3072,0.0074177781740824384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,8192,0.024339555038346186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,12288,0.014464888307783337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,2560,0.006274666637182236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,3072,0.013831111292044321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,2560,0.0069884442620807225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,2560,0.007637333538797166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,3584,0.008370666868156856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,2048,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,2048,0.01146488885084788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,1536,0.004936000125275718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,3072,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,1536,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,2560,0.01275022245115704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,1024,0.004052444464630551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,1536,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,8192,0.011408888631396823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,1024,0.00942844483587477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,768,0.0039013334446483185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,1024,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,2048,0.005487111293607288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,1536,0.01034755590889189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,512,0.003721777763631609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,768,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,1024,0.005756444401211209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,2048,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,256,0.003359111025929451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,512,0.006264888991912206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,768,0.005647999958859549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,768,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,128,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,256,0.00601333338353369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,128,0.004943111290534337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,512,0.008663111262851292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,64,0.0029120000286234748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,128,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,32,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,256,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,256,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,768,65536,0.030600888861550227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,65536,0.06218844652175903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,3584,0.007640000018808577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,65536,0.14440533849928114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,65536,0.057479110029008657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,51200,0.04906222224235535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,64,0.004909333255555895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,32,0.0052879999081293745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,51200,0.1143946647644043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,51200,0.046201778782738574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,16384,0.01685688893000285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,768,12288,0.01311644415060679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,512,0.005274666680230035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,16384,0.04187822341918945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,128,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,768,10240,0.01182577759027481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,12288,0.014169777433077494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,12288,0.015619556109110514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,12288,0.0329111119111379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,10240,0.01312711089849472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,10240,0.01383733335468504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,768,16384,0.013447999954223633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,8192,0.011390222443474663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,768,7168,0.00981866651111179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,8192,0.024170666933059692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,5120,0.00942666663063897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,768,51200,0.026294221480687458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,7168,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,7168,0.022384888596004907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,768,8192,0.011006222003036074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,6144,0.010036444498433007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,10240,0.02864355511135525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,6144,0.010038221875826517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,6144,0.019765333996878732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,8192,0.011924444801277585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,768,4096,0.008091555701361762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,16384,0.019122666782803006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,5120,0.017909333109855652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,768,6144,0.009350222018029954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,4096,0.00868088917599784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,4096,0.015736889508035447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,3584,0.007663999994595845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,3584,0.014835556348164877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,768,3072,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,3584,0.0083155557513237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,5120,0.009411555197503831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,3072,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,7168,0.011052444577217102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,768,3584,0.007287999822033777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,3072,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,2560,0.006999111009968652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,2560,0.012794666820102267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,768,2048,0.005562666803598404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,5120,0.00925511121749878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,2048,0.006643555644485686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,4096,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,768,2560,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,3072,0.013965333501497904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,2048,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,1536,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,768,5120,0.008663111262851292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,768,1024,0.004218666503826777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,1536,0.010420444111029306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,2560,0.007635555333561367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,768,768,0.003879111260175705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,1024,0.006646222124497096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,1024,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,1024,0.009466666314336989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,768,512,0.0035582222044467926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,768,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,768,1536,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,768,0.009001777403884465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,768,256,0.003269333392381668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,512,0.006036444256703059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,512,0.008671999805503422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,256,0.005193777796294955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,768,128,0.00318755561278926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,256,0.007736889024575551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,128,0.005974222388532426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,128,0.004964444372389051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,768,64,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,768,32,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,2048,0.011727111207114326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,64,0.004993777722120285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,512,65536,0.025386666258176167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,512,0.005283555636803309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,1536,0.00609422226746877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,65536,0.05720711416668362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,65536,0.06102844741609362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,65536,0.14421155717637804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,128,0.007663110891977946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,51200,0.04535999894142151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,512,16384,0.011900444825490316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,51200,0.11414666970570882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,16384,0.016920000314712524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,32,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,512,12288,0.013825777504179211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,16384,0.04126311010784573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,256,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,512,51200,0.01977066695690155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,12288,0.014176888598336114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,12288,0.015166223049163818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,12288,0.03259911139806112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,512,8192,0.010494222243626913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,10240,0.013462222284740873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,768,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,10240,0.02827733423974779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,512,7168,0.009702222214804756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,8192,0.011408000356621213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,8192,0.01183644433816274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,16384,0.01864177816443973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,512,6144,0.009395555489593083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,7168,0.02200444373819563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,6144,0.010098666780524785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,10240,0.013000888956917657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,51200,0.04832266767819723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,6144,0.010070222119490305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,5120,0.009104889300134447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,8192,0.02405511173937056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,7168,0.011053333679835001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,5120,0.009347555538018545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,512,10240,0.012073777616024017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,4096,0.008228444390826756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,512,3584,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,4096,0.01586666703224182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,512,5120,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,6144,0.019722667005327012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,512,3072,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,3584,0.00832177781396442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,3584,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,3584,0.014570666684044732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,512,2560,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,3072,0.007972444097201029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,512,4096,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,3072,0.013800889253616333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,2560,0.007692444655630324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,2560,0.007000000112586551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,4096,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,2048,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,512,1536,0.0047973332305749255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,5120,0.01800355480776893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,1536,0.006110222389300664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,7168,0.010781333678298526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,512,1024,0.003932444585694207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,1536,0.0069857777820693115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,512,2048,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,1536,0.010418666733635796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,2048,0.006646222124497096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,1024,0.005717333406209946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,1024,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,768,0.00565244464410676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,768,0.008955555657545725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,512,512,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,2560,0.012759110993809171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,512,0.0052791109515561005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,1024,0.006648889018429651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,512,256,0.003165333428316646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,512,0.005995555470387141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,512,0.008657777474986183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,256,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,3072,0.007295110987292395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,256,0.007638221813572778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,768,0.006326222171386083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,128,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,128,0.005984888722499211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,128,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,512,32,0.0026942222482628296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,512,768,0.003587555554178026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,64,0.005135111096832487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,32,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,256,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,256,51200,0.015440889530711703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,65536,0.05963377820120918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,65536,0.14374311765034994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,512,64,0.0030737777964936364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,51200,0.04048088855213589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,51200,0.1141244437959459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,2048,0.01143555591503779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,256,65536,0.018232888645595975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,16384,0.018245332770877414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,512,128,0.0028479999552170434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,65536,0.05226755473348829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,12288,0.014848000473446317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,256,10240,0.009311999711725447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,12288,0.014151111245155334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,12288,0.03257866700490316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,51200,0.04759555392795139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,256,16384,0.010738666686746808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,10240,0.012997333374288348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,10240,0.028135998381508723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,256,12288,0.009008888569143083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,256,7168,0.007982222570313348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,8192,0.011448000040319232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,7168,0.01106844428512785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,7168,0.021653332644038733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,7168,0.010728889041476779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,10240,0.013458666702111563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,6144,0.010032888915803697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,8192,0.02405422263675266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,6144,0.010027555955780877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,5120,0.009007111191749573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,256,6144,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,5120,0.009030222064918941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,256,4096,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,5120,0.017870222528775532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,16384,0.01687111126052009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,6144,0.019918221566412184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,4096,0.00868977771864997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,256,3584,0.0070017774899800616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,4096,0.015483554866578845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,3584,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,3584,0.00833155545923445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,256,3072,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,3072,0.0073520003093613525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,8192,0.011591999895042844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,3072,0.013784888717863293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,256,2560,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,3072,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,256,5120,0.00810222244924969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,4096,0.008038221961922115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,2560,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,2560,0.00693866651919153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,2560,0.012434666355450949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,256,1536,0.00453244439429707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,2048,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,3584,0.014653333359294467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,2048,0.011393778026103973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,256,1024,0.0038862224254343244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,1536,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,1536,0.005974222388532426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,1536,0.010286221901575724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,256,768,0.003564444267087512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,16384,0.040582223070992365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,1024,0.009691555466916826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,768,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,256,512,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,2048,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,512,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,512,0.008692444198661381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,512,0.00601333338353369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,1024,0.005647999958859549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,256,2048,0.005056000004212062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,768,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,256,128,0.002867555452717675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,768,0.00904799997806549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,256,8192,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,128,0.005044444567627377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,128,0.008058666355080074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,1024,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,256,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,256,0.00798044436507755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,32,0.004976888911591636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,128,65536,0.016062221593326993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,128,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,256,64,0.0028808888875775864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,65536,0.04968533251020643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,128,51200,0.013825777504179211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,256,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,65536,0.14301333162519667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,256,32,0.002567111204067866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,128,16384,0.009676444033781687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,51200,0.04003644320699904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,51200,0.046543111403783165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,51200,0.11349333657158746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,16384,0.017906667457686532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,128,12288,0.0084915558497111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,256,256,0.0032097777972618737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,16384,0.04058400127622817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,128,10240,0.00902133352226681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,12288,0.014852444330851236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,12288,0.0325546662012736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,64,0.004942222187916438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,128,8192,0.00835111074977451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,10240,0.013095999757448832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,10240,0.02804888950453864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,128,7168,0.00794488853878445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,8192,0.011429333024554782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,16384,0.016918222109476726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,7168,0.01071911139620675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,128,6144,0.0074720001882976955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,7168,0.021670222282409668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,6144,0.009722666607962714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,10240,0.012816000315878125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,128,5120,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,6144,0.01959822244114346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,6144,0.009758222434255812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,8192,0.011412444214026133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,128,4096,0.007355555891990662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,5120,0.009191110730171204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,5120,0.009005332986513773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,5120,0.01756000022093455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,4096,0.00811111099190182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,4096,0.015153777268197803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,3584,0.007698666718271043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,3584,0.014488889111412896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,3584,0.008292444050312042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,8192,0.0240328891409768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,7168,0.0107351111041175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,4096,0.008676444490750631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,128,2560,0.006434666613737742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,12288,0.014152000347773233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,3072,0.013783999615245394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,2560,0.007329777710967594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,2560,0.006955555743641323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,2560,0.012789333032237159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,128,3072,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,2048,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,128,1536,0.004562666846646203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,2048,0.011416888899273343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,3072,0.007326222128338284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,65536,0.0571297804514567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,128,1024,0.0038951109680864546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,1536,0.010438222024175854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,128,2048,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,1024,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,128,768,0.003583999971548716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,1024,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,1024,0.009689778089523315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,2048,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,768,0.005312888986534542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,768,0.008981333010726506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,128,3584,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,1536,0.007017778025733099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,128,256,0.0029973333908451926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,512,0.00869777798652649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,256,0.004998222407367495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,256,0.007948444121413762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,768,0.006314666734801398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,128,512,0.003208000005947219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,3072,0.008052444292439355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,1536,0.005998222364319696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,512,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,128,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,128,32,0.002507555608948072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,64,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,64,65536,0.014496889379289416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,32,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,256,0.006000888844331105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,64,51200,0.012420444852775998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,64,16384,0.009429333110650381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,65536,0.05679111348258125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,64,12288,0.00868000007337994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,51200,0.045336888896094434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,512,0.005969777703285217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,12288,0.014635556273990206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,10240,0.01314222233162986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,64,8192,0.007318221860461765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,128,128,0.002624888800912433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,64,7168,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,8192,0.011426666544543372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,64,6144,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,128,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,128,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,64,5120,0.006304889089531369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,7168,0.010755555497275459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,64,4096,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,64,3584,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,5120,0.009033777647548253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,64,3072,0.006329777754015393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,4096,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,64,2560,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,3072,0.007326222128338284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,64,2048,0.006334222025341458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,2560,0.006866666591829724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,64,1536,0.004387555436955558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,6144,0.009764444496896531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,64,1024,0.003641777982314428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,1536,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,64,768,0.003544888976547453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,3584,0.007650666766696506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,64,512,0.003220444545149803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,768,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,128,64,0.002846222163902389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,256,0.004968889057636261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,2048,0.006602666858169768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,64,64,0.002530666689078013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,128,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,64,32,0.002570666579736604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,1024,0.005623999983072281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,32,65536,0.013812444276279874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,64,256,0.0028728888266616394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,32,51200,0.012504888905419244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,64,10240,0.007695111135641734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,64,128,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,32,16384,0.009093333449628618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,51200,0.0452133317788442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,16384,0.017888888716697693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,32,10240,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,64,0.004960000101062986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,32,8192,0.007288888924651676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,32,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,10240,0.013443555268976422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,32,7168,0.0069573331210348345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,8192,0.011727111207114326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,32,6144,0.006496888895829518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,32,5120,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,32,12288,0.008369777765538957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,32,4096,0.005985777825117111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,6144,0.010098666780524785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,32,3584,0.006487111250559489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,5120,0.008785777621799046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,12288,0.014837332897716098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,32,3072,0.006236444330877728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,4096,0.008057778080304464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,32,2560,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,32,2048,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,3584,0.007695111135641734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,32,1536,0.004559999952713649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,2560,0.007009777757856581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,16384,0.01792711185084449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,32,1024,0.003903111235962974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,32,768,0.003572444534964032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,1536,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,32,512,0.0031928889867332247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,1024,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,32,256,0.002887111157178879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,512,0.005325333525737126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,32,128,0.002887111157178879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,256,0.005137777576843898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,32,64,0.0025235555238193935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,512,0.005280888742870755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,65536,0.05616088708241781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,256,32,32,0.0025742221623659134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,7168,0.010816888676749336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,32,0.004761777818202972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,768,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,3072,0.007334222396214803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,16384,0.4942195680406358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,128,0.005006222261322869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,64,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,51200,1.0797217686971028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,51200,1.5245822270711262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,16384,0.1432266632715861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,10240,0.24665954377916124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,2048,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,10240,0.22273333867390951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,8192,0.2067155573103163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,12288,0.3116062217288547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,10240,0.08942755725648667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,7168,0.1806551085578071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,12288,0.2601377699110243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,8192,0.07414399915271334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,6144,0.1536124414867825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,7168,0.15226399898529053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,7168,0.06513955858018664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,5120,0.12954310576121011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,6144,0.0569724440574646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,16384,0.3420071072048611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,5120,0.11010755432976617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,12288,0.11002933316760594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,4096,0.09022400114271377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,3584,0.09307022227181329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,6144,0.13499377833472356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,3584,0.03659644391801622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,3072,0.08492355876498753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,3584,0.07840089003245036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,51200,0.43091999159918887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,3072,0.03288266725010342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,3072,0.0684773325920105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,2560,0.02894044584698147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,2560,0.059398220645056836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,2048,0.06050311194525825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,5120,0.04846044381459554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,2048,0.0500311123000251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,1536,0.04641955428653293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,8192,0.17671199639638266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,1024,0.033139556646347046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,1536,0.019870221614837646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,2560,0.07463555865817599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,1024,0.01577333278126187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,1024,0.030834668212466772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,512,0.019211555520693462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,768,0.025106666816605464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,512,0.013066666821638743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,512,0.0214675549003813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,256,0.012801777985360889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,256,0.012085333466529846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,128,0.011343110766675739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,256,0.019039111004935372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,1536,0.03942044576009115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,64,0.009301332963837517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,128,0.011790222591824003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,768,0.014141332772043018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,32,0.009719111025333405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,64,0.017774222625626456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,32,0.01777333352300856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,4096,0.10590489043129815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,768,0.025839999318122864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,128,0.018234666850831773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,65536,1.3283057742648654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,51200,1.0257039599948459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,2048,0.025302223033375208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,16384,0.3378569020165338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,65536,0.42066311836242676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,4096,0.041780445310804576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,16384,0.11270311143663193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,16384,0.2720613214704725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,10240,0.17875199847751191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,51200,0.338864008585612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,12288,0.08699288633134629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,65536,1.0855564541286893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,10240,0.07125511434343126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,8192,0.14457333087921143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,7168,0.14237688647376165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,51200,0.8355502022637261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,6144,0.10882755782869126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,7168,0.05219911204444038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,7168,0.12257866064707439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,12288,0.2058631049262153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,6144,0.0458586679564582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,8192,0.16574311256408691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,4096,0.08232533269458346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,5120,0.039220445685916476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,5120,0.08782222535875107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,8192,0.058372444576687284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,12288,0.2539199988047282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,4096,0.033640000555250384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,4096,0.07224977678722806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,3072,0.06598399745093451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,6144,0.10269066360261704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,10240,0.16714133156670463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,3072,0.027076444692081873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,3072,0.05517066518465678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,2560,0.04900177650981479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,2048,0.046366221374935575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,3584,0.06299289067586263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,3584,0.06996622350480822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,1536,0.032384001546435885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,2048,0.020340444313155282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,2048,0.04068888889418708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,1536,0.016711999972661335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,1536,0.032455109887652926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,768,0.01862844493654039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,1024,0.013842666314707862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,1024,0.025260445144441392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,5120,0.09708799918492635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,512,0.014239999983045789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,768,0.012831110921170978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,256,0.012069332930776807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,512,0.011916444533401065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,512,0.017823111679818895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,128,0.010009778042634329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,256,0.015923556354310777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,1024,0.025282666087150574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,64,0.008727110922336578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,128,0.010728889041476779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,128,0.015169777803950839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,32,0.008986666798591614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,64,0.014881778094503613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,768,0.02066044509410858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,2560,0.02397777802414364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,65536,0.39120088683234316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,65536,0.36170045534769696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,65536,0.14538933171166316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,32,0.015146666102939181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,51200,0.2851262092590332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,51200,0.11559732755025227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,3584,0.02996088729964362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,2560,0.04787022206518385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,16384,0.040765331851111516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,12288,0.08484177456961738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,16384,0.09210399786631267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,10240,0.0706311133172777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,12288,0.03264088763131036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,51200,0.3084542221493191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,8192,0.06160089042451647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,10240,0.028120001157124836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,8192,0.023743111226293776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,8192,0.04941511154174805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,7168,0.021459556288189355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,256,0.010617777705192566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,7168,0.04304888844490051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,16384,0.10933599869410197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,6144,0.03808533483081394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,6144,0.018986667195955913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,5120,0.03268266717592875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,5120,0.017160000072585214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,10240,0.05916621949937609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,4096,0.015237333046065437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,3584,0.028918223248587713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,6144,0.04479111234347025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,3072,0.02548888822396596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,3584,0.013960889644092984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,3584,0.024667556087176006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,2560,0.021691555778185528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,3072,0.013008889224794177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,12288,0.06975466675228544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,2048,0.01815733313560486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,2560,0.01925511161486308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,7168,0.05263288815816244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,1536,0.014307555225160388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,2048,0.016593777471118502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,1536,0.009008888569143083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,5120,0.038704001241260104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,1536,0.014108444253603617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,1024,0.008043555749787224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,2048,0.011038222246699862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,4096,0.027727999620967444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,768,0.007633777956167857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,1024,0.01142133358452055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,1024,0.011745777395036487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,512,0.008674666285514832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,256,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,512,0.007335111498832703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,768,0.009362666971153682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,128,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,256,0.006991110742092133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,256,0.008218666745556725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,64,0.004909333255555895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,768,0.010141332944234213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,512,0.007624000310897827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,128,0.00832444429397583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,64,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,32,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,65536,0.29597510231865776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,51200,0.23577422565884062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,65536,0.2907493379380968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,65536,0.12876089413960776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,16384,0.08849155240588719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,128,0.0070159998204973005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,32,0.005311111195219888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,51200,0.10196355316374038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,51200,0.22601599163479277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,16384,0.07651378048790826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,2560,0.012101333174440594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,10240,0.055008888244628906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,12288,0.030216889248953924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,12288,0.0589617755677965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,8192,0.04574577675925361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,10240,0.02589688864019182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,7168,0.04084622197681003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,8192,0.022208889325459797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,8192,0.04206755426194933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,7168,0.01997866729895274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,7168,0.037010666396882795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,12288,0.06711733341217041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,16384,0.03849422269397312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,5120,0.029438220792346533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,3072,0.021735999319288466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,4096,0.024513777759340074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,5120,0.016340444485346477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,5120,0.02796088986926609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,3584,0.021458667185571458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,4096,0.01458044515715705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,10240,0.04983911249372694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,4096,0.023418666587935552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,3072,0.01917155583699544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,3584,0.021067554752031963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,3072,0.012785777449607849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,6144,0.018208000395033095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,2560,0.01627200014061398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,4096,0.032491554816563926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,2048,0.013802666631009845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,2560,0.011416888899273343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,2560,0.016864889197879367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,1536,0.012348444097571902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,2048,0.010048888623714447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,2048,0.014846222268210517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,1024,0.009676444033781687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,1536,0.008616888688670265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,6144,0.033258666594823204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,768,0.00833155545923445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,1024,0.007854222423500484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,1024,0.010128888818952773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,512,0.006656000183688269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,6144,0.03264710969395108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,768,0.008674666285514832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,256,0.005249777601824866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,512,0.007303111255168915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,512,0.008011555506123437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,128,0.004265777766704559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,256,0.007670222057236566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,1536,0.01275111072593265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,128,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,3072,0.019144889381196763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,768,0.007663110891977946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,64,0.007356444166766272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,65536,0.23027377658420137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,32,0.007324444750944774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,256,0.0069982219073507525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,51200,0.21897510687510172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,65536,0.24718133608500162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,65536,0.12395289209153916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,16384,0.0629440016216702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,51200,0.0997048881318834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,16384,0.03709244396951463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,16384,0.06595200300216675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,12288,0.028759111960728962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,10240,0.04376622372203403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,3584,0.013462222284740873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,128,0.006979555719428592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,10240,0.04284177886115181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,51200,0.1929324467976888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,32,0.003923555629120933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,7168,0.03129333257675171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,8192,0.021427555216683283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,12288,0.050388445456822716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,6144,0.027393778165181477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,64,0.00388711111413108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,6144,0.028176890479193792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,5120,0.0233297778500451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,6144,0.017668444249365065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,12288,0.052397333913379245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,4096,0.019860444797409903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,5120,0.015998222761683993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,5120,0.023992000354660883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,3584,0.017229333519935608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,4096,0.014243554737832812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,4096,0.02049688829316033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,3072,0.015489778584904142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,3584,0.01861422260602315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,10240,0.025098666548728943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,8192,0.03580177823702494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,3072,0.012460444536474017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,2560,0.010842666857772403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,8192,0.036426666710111834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,7168,0.01955111159218682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,7168,0.031940443648232356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,2048,0.013164444102181328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,3584,0.01309866706530253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,1536,0.008456000023418004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,2560,0.013452444639470843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,1536,0.011395555403497485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,1024,0.007693332930405934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,768,0.006728888799746831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,1024,0.009348444640636444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,768,0.007330666813585494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,512,0.005583111196756363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,768,0.008107555409272512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,1536,0.010128888818952773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,256,0.004559999952713649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,512,0.00699733363257514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,512,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,1024,0.00814666681819492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,128,0.004222222086456087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,256,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,64,0.003933333274390963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,128,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,32,0.004242666479614046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,64,0.0069546666410234236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,32,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,2048,0.009362666971153682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,65536,0.24039822154574922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,3072,0.017049777838918898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,51200,0.18727555539872912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,2560,0.015223999818166098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,51200,0.1585306723912557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,16384,0.06411555740568373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,16384,0.054085334142049156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,256,0.0069857777820693115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,65536,0.07863555351893108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,12288,0.052728000614378184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,16384,0.02437688906987508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,2048,0.011787555283970304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,12288,0.01959733333852556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,12288,0.0422382222281562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,10240,0.017242666747834947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,10240,0.03614133265283372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,8192,0.015209777487648858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,7168,0.031210667557186548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,65536,0.2022648917304145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,7168,0.014136888914638095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,7168,0.026781333817376032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,8192,0.03563822309176127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,5120,0.02366755571630266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,6144,0.01314399970902337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,8192,0.030324445830451116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,5120,0.01181688904762268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,5120,0.020578665865792167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,4096,0.010829333629873065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,3584,0.01724799970785777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,4096,0.017677333619859483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,51200,0.06186222367816501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,3584,0.010064000056849586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,3072,0.01548266741964552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,3584,0.016237333416938782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,2560,0.013435555828942193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,3072,0.009026666482289633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,3072,0.014874666929244995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,4096,0.01994311147265964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,2560,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,2560,0.013432000246312885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,1536,0.009929777847396003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,6144,0.02755733331044515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,1536,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,1024,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,6144,0.023437332775857713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,1024,0.0070062221752272705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,1024,0.008325333396593729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,2048,0.011782222323947482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,128,0.006959111326270633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,512,0.0052435555391841466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,1536,0.009813333551088968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,512,0.00700711127784517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,256,0.004229333251714706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,512,0.006331555545330048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,768,0.005919111271699269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,128,0.003911111089918348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,256,0.006613333192136552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,10240,0.04381422201792399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,128,0.006353777729802662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,768,0.007439999944633908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,128,0.006324444380071428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,64,0.006568000134494569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,32,0.006728888799746831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,65536,0.1860506667031182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,256,0.006302222195598815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,51200,0.14045688841078016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,65536,0.19739911291334364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,2048,0.011979555918110741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,32,0.004200889004601372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,16384,0.05243911014662849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,51200,0.059340443876054555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,51200,0.15437866581810847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,16384,0.023342221975326538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,16384,0.05320088730918037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,10240,0.03436355458365546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,12288,0.01869333287080129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,12288,0.041138665543662176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,10240,0.016520000166363187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,65536,0.07435555590523614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,2048,0.007978666987684038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,7168,0.024961777859263953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,8192,0.01464355488618215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,7168,0.013767110804716745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,7168,0.026100445124838088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,6144,0.022188444932301838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,6144,0.012515555653307172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,5120,0.018761777215533786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,6144,0.023197333017985027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,8192,0.02793244520823161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,10240,0.03524444500605265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,5120,0.011629333098729452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,64,0.00388711111413108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,3584,0.014422222971916199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,4096,0.010388444695207808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,4096,0.017447999782032438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,3072,0.012746666868527731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,3584,0.009341333475377824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,3584,0.015853333804342482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,2560,0.011691555380821228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,3072,0.014626666903495789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,2560,0.008008889026112026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,2048,0.010042666561073726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,5120,0.020219556159443326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,8192,0.029603554142846003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,2048,0.011402666568756104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,1536,0.008347555167145198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,3072,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,1536,0.009354666703277165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,1024,0.006761777732107375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,2560,0.0131217779384719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,1024,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,768,0.005889777921968036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,2048,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,768,0.006604444649484422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,768,0.007359999749395583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,1536,0.00720977783203125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,4096,0.015820443630218506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,512,0.0069520001610120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,256,0.004259555704063839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,1024,0.006741333338949416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,256,0.0064586665895250105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,12288,0.04086577892303467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,128,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,512,0.0052435555391841466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,128,0.006274666637182236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,32,0.003535110917356279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,768,0.006702222343948152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,512,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,64,0.0063288890653186375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,51200,0.13388355573018393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,65536,0.07050489054785834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,128,0.004003555410438114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,65536,0.15874132845136854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,51200,0.05748266643948025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,51200,0.12422222561306423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,64,0.003632889025741153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,12288,0.03810311026043362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,16384,0.022169777088695105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,10240,0.031160887744691636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,12288,0.018160889546076458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,65536,0.1913155582216051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,32,0.007258666886223688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,10240,0.016139555308553908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,256,0.006311111152172089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,7168,0.02239199976126353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,8192,0.024479111035664875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,7168,0.013339555925793119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,7168,0.02169599963559045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,16384,0.04346222347683377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,6144,0.012337778177526263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,8192,0.026329777306980554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,5120,0.017321777012613084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,10240,0.028919110695521038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,4096,0.014936000108718872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,5120,0.017051556044154696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,4096,0.009526222116417354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,4096,0.014935111006100973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,3584,0.013821333646774292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,3584,0.008848889006508721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,3072,0.011745777395036487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,16384,0.04804089003139072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,3072,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,6144,0.01973600023322635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,2560,0.011048888994587792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,2560,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,2048,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,8192,0.014293332894643148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,2048,0.009758222434255812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,1536,0.008316444853941599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,2048,0.007373332977294922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,3072,0.01277599980433782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,3584,0.013206221991115145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,1024,0.006697777658700943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,1536,0.008367111285527548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,768,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,1024,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,6144,0.019542222221692402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,512,0.004601777841647466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,768,0.006576889091067844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,768,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,256,0.003912888881233004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,512,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,512,0.006520888871616787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,128,0.0035955554081334006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,256,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,1536,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,64,0.0035013332962989807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,128,0.006130666782458623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,128,0.006189333481921091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,32,0.0035377778112888336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,64,0.005972444597217772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,32,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,256,0.006258666515350342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,65536,0.15448799398210314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,51200,0.12624977694617376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,51200,0.05500266618198819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,16384,0.0429057776927948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,65536,0.16033600436316595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,5120,0.01126933346192042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,12288,0.03288799855444167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,16384,0.04225155711174011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,65536,0.06823911269505818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,12288,0.01756622228357527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,12288,0.03251377741495768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,2560,0.011545778148704104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,10240,0.028021332290437486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,8192,0.023391111029518977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,16384,0.021176000436147053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,8192,0.013956444131003486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,7168,0.02096533278624217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,10240,0.027601778507232666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,7168,0.01292711082432005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,6144,0.018198221921920776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,1024,0.007351111206743453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,10240,0.01573600040541755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,5120,0.015769778026474845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,8192,0.0237137774626414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,5120,0.016872889465755887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,4096,0.01346577786737018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,5120,0.01053955571519004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,4096,0.009047110875447592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,3584,0.0127688886390792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,6144,0.011894221935007306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,6144,0.01920977731545766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,3072,0.011384000380833944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,3584,0.013682666752073499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,3072,0.008159110943476358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,3072,0.012431110772821637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,7168,0.02141155633661482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,2048,0.008906666603353288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,2560,0.010791999598344168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,1536,0.00701155596309238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,2048,0.007355555891990662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,2048,0.009367999931176504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,3584,0.00866844422287411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,1024,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,1536,0.007017778025733099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,768,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,1024,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,1024,0.007293333609898885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,512,0.004551110996140374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,768,0.006575111299753189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,51200,0.12081244256761338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,4096,0.014820444915029736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,2560,0.007697777615653143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,256,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,128,0.003640888879696528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,1536,0.008345777789751688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,128,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,64,0.0035911111368073356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,768,0.006704000135262807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,32,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,512,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,512,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,64,0.005972444597217772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,65536,0.13166844844818115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,32,0.006240889016124938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,51200,0.10462844371795654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,65536,0.06594755252202351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,65536,0.1144675546222263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,16384,0.04071733355522156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,51200,0.05249511202176412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,51200,0.090137779712677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,12288,0.029947555727428857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,16384,0.020308444897333782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,10240,0.025655999779701233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,12288,0.017156443662113614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,256,0.0038968887594011095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,10240,0.015196444259749519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,10240,0.022030222747060988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,256,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,8192,0.013471110827393003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,2560,0.010372444159454769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,128,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,7168,0.017311111092567444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,16384,0.03219377663400438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,7168,0.012248000337017907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,5120,0.01425155500570933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,12288,0.02539377742343479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,8192,0.020968000094095867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,6144,0.015612444943851896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,4096,0.0127688886390792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,5120,0.013802666631009845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,3584,0.011741333537631564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,4096,0.012272000312805176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,3584,0.008424888882372115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,6144,0.016390222642156813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,3072,0.01072799993885888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,6144,0.011063999599880643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,2560,0.010021333065297868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,5120,0.009476444787449306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,2560,0.007838221887747446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,2560,0.009009777671760982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,3584,0.010949333508809408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,2048,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,1536,0.007184888753626082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,2048,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,3072,0.008040889269775814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,1024,0.005712000032265981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,1536,0.006959999601046245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,1536,0.007296000089910295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,768,0.005241777747869492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,1024,0.006637333167923822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,4096,0.008742222355471717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,2048,0.008293333152929941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,512,0.004560889055331548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,768,0.00630044440428416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,768,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,256,0.003938666648334927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,512,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,256,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,3072,0.010029333333174387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,128,0.003556444413132138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,256,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,64,0.003555555724435382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,128,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,7168,0.018561777141359117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,32,0.0036551112102137674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,128,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,65536,0.1002275546391805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,64,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,32,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,512,0.006153777655627992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,65536,0.11640177832709418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,16384,0.033857779370413885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,51200,0.0515342222319709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,51200,0.08954222334755792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,16384,0.020261334048377145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,16384,0.031425777408811785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,8192,0.019003555178642273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,12288,0.01684888866212633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,12288,0.025252444876564875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,8192,0.020051555501090158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,10240,0.015127999915017022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,10240,0.021868444151348535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,51200,0.08631377749972874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,7168,0.017870222528775532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,8192,0.0188711103465822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,6144,0.015839111473825242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,7168,0.012048888537618848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,12288,0.028913779391182795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,5120,0.013789333403110504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,6144,0.011013333168294696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,6144,0.015662221444977652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,10240,0.024115555816226538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,5120,0.01350044459104538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,4096,0.008730666504965888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,8192,0.013083555632167392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,7168,0.017052445146772597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,3584,0.01071200023094813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,3072,0.009008000294367472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,3584,0.00835111074977451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,4096,0.01106844428512785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,2560,0.008264888491895463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,3072,0.009688888986905416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,4096,0.011734222372372946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,65536,0.06463466750250922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,2048,0.008202667037645975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,2048,0.007344889144102733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,1536,0.0063946665161185795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,2560,0.007696888513035244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,1536,0.007335999773608313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,1024,0.005439110928111606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,1536,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,1024,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,768,0.004959110998445087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,1024,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,3072,0.008019555773999956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,512,0.004218666503826777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,768,0.006239111224810283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,512,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,512,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,256,0.005566222386227713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,5120,0.009696889254781935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,256,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,128,0.0032257777121331957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,768,0.006340444501903322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,128,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,128,0.005588444570700328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,3584,0.010204444328943888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,32,0.0032035555276605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,64,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,32,0.005632888939645555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,1024,0.006637333167923822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,51200,0.0755022234386868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,65536,0.10970311032401191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,65536,0.06339822212855022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,16384,0.03256888853179084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,51200,0.05044088761011759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,51200,0.08540088600582546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,2560,0.008792888787057664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,12288,0.02200177808602651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,16384,0.03035911255412632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,10240,0.019595555133289762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,65536,0.09892800119188096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,12288,0.016573333077960543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,8192,0.016141333513789706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,10240,0.014868444866604276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,10240,0.02113955550723606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,7168,0.014743111199802823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,8192,0.018290667070282828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,64,0.00327822214199437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,7168,0.01112711098459032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,7168,0.016584000653690763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,2048,0.0070053330726093715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,256,0.0038862224254343244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,5120,0.012111999922328524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,6144,0.010383110907342698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,6144,0.0151591118839052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,12288,0.02465244465404087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,5120,0.009593778186374241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,5120,0.013046222428480784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,3584,0.009723555710580613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,8192,0.012450666891203986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,4096,0.011179555621412067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,3072,0.008839999636014303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,3584,0.008351999852392409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,3584,0.010192888478438059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,6144,0.013212444053755866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,2560,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,3072,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,2048,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,2560,0.007650666766696506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,2560,0.008719111482302347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,4096,0.010388444695207808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,2048,0.008022222254011366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,4096,0.00869511150651508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,1536,0.006980444822046492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,1024,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,1024,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,768,0.004260444392760594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,1024,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,16384,0.020064888728989493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,3072,0.009530666801664565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,768,0.006326222171386083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,768,0.005985777825117111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,1536,0.005395555661784277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,512,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,256,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,1536,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,128,0.003228444399105178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,64,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,128,0.006020444548792309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,32,0.0028808888875775864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,128,0.0053031109273433685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,512,0.0038648889296584656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,64,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,32,0.005704888867007361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,2048,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,65536,0.062012447251213916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,51200,0.06892622179455228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,16384,0.028528889020284016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,51200,0.04946133494377136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,256,0.005588444570700328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,51200,0.08375555276870728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,12288,0.022263111339675054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,16384,0.01974933346112569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,16384,0.029887109994888306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,12288,0.016361777981122334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,12288,0.02404444416364034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,65536,0.08527911371654934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,10240,0.014474666780895658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,10240,0.020899555749363367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,256,0.0035502223504914176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,7168,0.01352266636159685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,65536,0.1064373321003384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,8192,0.017877333694034152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,6144,0.012129777835475074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,7168,0.010828444527255164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,5120,0.01109333336353302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,6144,0.010306666294733683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,5120,0.012468444804350534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,4096,0.010388444695207808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,8192,0.015011555618709989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,4096,0.00868800034125646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,8192,0.012251555919647217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,3584,0.009321778184837764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,4096,0.010817777779367236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,3072,0.008857777549160851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,7168,0.01625066664483812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,3584,0.010056888891590966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,2560,0.007970666719807519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,3072,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,10240,0.023339556323157415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,3072,0.009413333402739631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,2048,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,2560,0.0076657773719893555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,5120,0.009356444080670675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,2560,0.008736888567606607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,2048,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,1024,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,2048,0.008037333687146505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,512,0.006158222340875202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,1536,0.007254222200976477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,1024,0.006666666517655055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,1024,0.006527110934257507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,3584,0.008347555167145198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,768,0.0063457778758472866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,768,0.006000888844331105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,256,0.003581333491537306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,6144,0.01443377799457974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,512,0.00602311102880372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,512,0.005711110929648082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,128,0.003216888962520493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,1536,0.006265777680608962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,256,0.005619555711746216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,64,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,1536,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,32,0.0029928889125585556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,768,0.004566222015354368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,65536,0.06985866361194186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,64,0.005388444496525659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,512,0.0041946665280395085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,32,0.005672889037264719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,51200,0.056702223089006215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,65536,0.06041600306828817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,16384,0.024205333656734888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,51200,0.048008888959884644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,256,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,51200,0.052829331821865506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,12288,0.020319110817379422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,16384,0.019495111372735765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,128,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,10240,0.019256888164414298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,12288,0.0155404441886478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,8192,0.014575999644067554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,10240,0.013179555535316467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,10240,0.014929778046078153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,7168,0.013250666360060373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,8192,0.011517333487669626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,8192,0.01292799992693795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,6144,0.012303111453851065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,7168,0.010864888628323873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,5120,0.01036711119943195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,7168,0.011747555600272285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,6144,0.010101333260536194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,6144,0.010375111467308467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,5120,0.009339555270142024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,16384,0.019987555013762582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,5120,0.009354666703277165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,3584,0.008772444393899705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,4096,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,3072,0.007625777688291337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,4096,0.008360000120268928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,3584,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,2560,0.006688000013430913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,3072,0.007646222081449296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,65536,0.06684088706970215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,2048,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,2560,0.007639110916190677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,2560,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,1536,0.005466666486528184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,4096,0.009256000320116678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,2048,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,1024,0.004567111117972268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,2048,0.006600889066855113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,1536,0.0069573331210348345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,12288,0.016639111770523917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,768,0.003968888686762916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,1024,0.006637333167923822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,1024,0.005963555640644497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,768,0.006334222025341458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,3072,0.008031999899281396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,768,0.005673777725961473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,256,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,512,0.006017777654859755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,512,0.0053688887920644544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,256,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,256,0.005140444470776452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,128,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,128,0.004967111266321606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,32,0.0032479998966058097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,1536,0.006254222244024277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,3584,0.008344000412358178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,65536,0.0554328891966078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,32,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,65536,0.05888977977964613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,65536,0.06428177489174737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,128,0.0033537778589460584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,16384,0.018960000740157235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,51200,0.04684799909591675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,51200,0.05061066812939114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,12288,0.016252444850073922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,16384,0.018586667047606576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,16384,0.01978933314482371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,64,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,12288,0.014190221826235453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,512,0.0038986665507157645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,12288,0.016527111331621807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,8192,0.01406044430202908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,10240,0.013054221868515015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,51200,0.04479733440611097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,8192,0.011593778100278644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,8192,0.012436444560686747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,6144,0.011384888655609555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,7168,0.011212444139851464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,128,0.005250666704442766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,6144,0.010060444474220276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,10240,0.01723288827472263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,5120,0.00905600024594201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,4096,0.008968888885445064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,7168,0.010783111055692038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,4096,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,4096,0.008658666577604082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,6144,0.010407999985747868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,3584,0.00831822223133511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,3584,0.008310221963458592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,3072,0.00739822205570009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,7168,0.012628444366984896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,3072,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,3072,0.007375111182530721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,64,0.003240888938307762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,5120,0.010298666854699453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,2560,0.0070186663005087115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,5120,0.009369778136412302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,2048,0.006610666712125142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,1536,0.005587555468082428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,2048,0.007340444458855524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,2560,0.007345777418878343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,1536,0.0069857777820693115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,1024,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,2560,0.007687999970383114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,2048,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,768,0.004287111262480418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,3584,0.007961778177155389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,512,0.0038968887594011095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,768,0.005611555443869696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,256,0.0035333335399627686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,512,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,512,0.0052871112194326185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,128,0.0032142222755485107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,1536,0.006269333263238271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,128,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,64,0.002867555452717675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,1024,0.005715555614895291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,32,0.0028942223224374983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,768,0.0063546668324205614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,10240,0.014494222071435718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,65536,0.03974399964014689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,32,0.0053022222386466126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,51200,0.032446222172843084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,65536,0.05760444535149468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,256,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,256,0.004942222187916438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,51200,0.04627289043532478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,1024,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,51200,0.04853777752982246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,12288,0.014153778553009033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,16384,0.016913778252071805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,10240,0.0131217779384719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,12288,0.014180445008807711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,64,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,8192,0.01219733307758967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,10240,0.013651555611027611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,65536,0.06147199869155884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,16384,0.015501333607567681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,128,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,16384,0.019295111298561096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,7168,0.010967111421955956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,12288,0.01569422251648373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,10240,0.012827555338541666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,6144,0.010058666268984476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,6144,0.010094222095277574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,8192,0.01145600030819575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,7168,0.01146577795346578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,5120,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,4096,0.00868800034125646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,3584,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,7168,0.010716444088353051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,3584,0.00833155545923445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,3072,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,3584,0.00775911079512702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,3072,0.008011555506123437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,5120,0.009401777552233802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,4096,0.007974222302436829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,2560,0.007634667058785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,2048,0.005588444570700328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,6144,0.010718222293588849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,2048,0.007294221884674496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,1536,0.004935111022657818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,5120,0.009703999592198266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,1536,0.0069546666410234236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,8192,0.011863999896579318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,1024,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,4096,0.008041777544551426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,768,0.0038764443662431505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,1024,0.006660444455014334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,2048,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,768,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,512,0.003342222215400802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,768,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,256,0.0033902221669753394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,512,0.00619911112719112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,2560,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,512,0.005295111073387994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,256,0.005995555470387141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,256,0.004997333304749595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,64,0.0028826666788922418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,2560,0.00701422244310379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,128,0.004953777624501122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,32,0.002883555574549569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,768,65536,0.032030221488740705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,64,0.004935111022657818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,32,0.004955555415815777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,65536,0.05695733096864489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,65536,0.05999288956324259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,128,0.002879111096262932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,51200,0.04610755708482531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,51200,0.04754044281111824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,16384,0.01848355597919888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,16384,0.016872889465755887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,768,12288,0.013576000101036496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,128,0.00600000015563435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,768,16384,0.013798221945762634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,12288,0.015189333094490899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,12288,0.014448000325096978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,10240,0.013105777402718862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,10240,0.013245333400037555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,3072,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,768,51200,0.025975111458036635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,8192,0.011730666789743634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,7168,0.010801778071456485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,768,10240,0.012708444562223224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,7168,0.01075644459989336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,6144,0.010046222143703038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,768,5120,0.009352888498041365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,6144,0.009988444546858469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,1024,0.004231111043029361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,5120,0.009383999639087254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,768,4096,0.007975999679830339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,5120,0.009032888544930352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,768,7168,0.010479999913109673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,4096,0.008504889077610439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,4096,0.008115555677149031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,768,3072,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,768,6144,0.010068444742096795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,3584,0.007691555553012424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,768,2560,0.005991999887757831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,3072,0.008018666671382057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,3072,0.007337777978844113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,768,8192,0.011064888702498542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,2560,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,2560,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,1536,0.006274666637182236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,2048,0.007332444190979004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,8192,0.011587555209795633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,768,1024,0.00423200014564726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,1536,0.006980444822046492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,1536,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,768,768,0.0038657776183552216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,1024,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,1024,0.005633777628342311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,768,512,0.003536888708670934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,768,0.006321777900060018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,768,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,768,256,0.0031848889258172777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,512,0.006112888869312074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,768,1536,0.004949333353175057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,768,128,0.0032035555276605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,256,0.006263111200597551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,768,3584,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,768,64,0.002879111096262932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,128,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,768,32,0.0028906667398081887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,64,0.004933333231343163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,512,65536,0.02641155487961239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,32,0.004969777746333016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,768,2048,0.0052728888889153796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,512,51200,0.02420444455411699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,512,0.0052933332820733385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,65536,0.05855821900897556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,256,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,51200,0.04496533340877957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,51200,0.0467564430501726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,512,12288,0.009699555734793345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,16384,0.016901334126790363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,16384,0.017987555927700467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,512,10240,0.012074666718641916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,12288,0.014208889669842191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,12288,0.014828445182906257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,512,8192,0.010453333457310995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,10240,0.013369777964221107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,65536,0.05688889159096611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,512,7168,0.00978311151266098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,8192,0.011740444435013665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,7168,0.01073688848151101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,512,6144,0.009023110899660323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,6144,0.010031110710567897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,6144,0.010046222143703038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,512,5120,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,8192,0.011405333048767514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,2048,0.006677333265542984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,512,4096,0.00797155582242542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,128,0.004908444566859139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,7168,0.010808000134097205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,512,3584,0.007344000041484833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,4096,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,4096,0.008204444415039485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,512,3072,0.006320000108745363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,3584,0.00832088871134652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,3584,0.007673777639865875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,3072,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,3072,0.0073626670572492815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,512,16384,0.011439111497667102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,5120,0.009354666703277165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,512,2048,0.0053013331360287136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,2048,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,10240,0.012995555169052549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,512,1536,0.00489777781897121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,2048,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,512,2560,0.005974222388532426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,512,1024,0.003943110919660992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,1536,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,1536,0.006270222365856171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,1024,0.006602666858169768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,2560,0.007668444679843054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,768,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,768,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,512,512,0.003531555748648114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,512,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,512,256,0.0031946665710873077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,256,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,3584,0.00833777752187517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,512,128,0.003220444545149803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,128,0.005986666513813867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,512,64,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,512,0.005250666704442766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,512,32,0.002891555635465516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,64,0.004907555464241239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,256,65536,0.020111110475328233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,512,768,0.0038782221575578055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,32,0.005092444519201915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,2560,0.006991110742092133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,256,51200,0.016723554995324876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,65536,0.050267553991741605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,256,16384,0.01038755559258991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,51200,0.04010222355524699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,128,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,256,12288,0.010075555907355415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,16384,0.01688799924320645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,16384,0.01756177842617035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,256,10240,0.009527111219035255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,12288,0.014762666490342883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,10240,0.012797333300113678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,1024,0.0058702222175068325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,256,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,65536,0.05751111110051473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,256,7168,0.00833422193924586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,8192,0.011338666909270816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,51200,0.04627289043532478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,8192,0.01144977741771274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,256,6144,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,7168,0.010737777584128909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,256,5120,0.009012444151772393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,6144,0.00980711148844825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,5120,0.009404444032245213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,5120,0.008797333472304875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,256,8192,0.008354666332403818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,4096,0.00866844422287411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,256,3584,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,4096,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,7168,0.010727110836240979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,256,3072,0.00656444455186526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,12288,0.014145778285132514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,3072,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,256,2560,0.005913777897755305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,5120,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,10240,0.01312355531586541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,256,4096,0.007690666450394525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,2560,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,2560,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,256,1536,0.004550222307443619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,2048,0.006653333289755716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,3584,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,256,1024,0.003920888735188378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,3584,0.007637333538797166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,1024,0.006618666566080517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,256,768,0.003879111260175705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,1024,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,256,2048,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,768,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,768,0.00534400012758043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,2048,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,256,256,0.0032088889016045462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,512,0.005991999887757831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,512,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,256,128,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,1536,0.00702133360836241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,6144,0.010054222411579555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,256,64,0.0028568889117903183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,128,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,256,32,0.0025475554996066624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,64,0.005018666800525454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,256,512,0.0032097777972618737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,128,65536,0.0181377778450648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,32,0.005133333305517833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,128,51200,0.015388444066047668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,65536,0.04942933387226529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,65536,0.05548444390296936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,128,16384,0.009721777505344814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,256,0.005219555563396878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,3072,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,128,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,16384,0.01670222149954902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,128,12288,0.009823111196359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,12288,0.014129777749379476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,128,10240,0.009359111388524374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,12288,0.014481777946154276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,10240,0.012772444221708508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,128,8192,0.008340444829728868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,1536,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,51200,0.04443022277620104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,128,7168,0.009389333426952362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,8192,0.011415999796655444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,8192,0.011384000380833944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,128,6144,0.008741333252853816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,7168,0.01071911139620675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,7168,0.01073333372672399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,128,5120,0.008650666309727563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,6144,0.009711110757456886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,5120,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,128,4096,0.007639110916190677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,16384,0.017551110850440133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,4096,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,128,3584,0.006952889263629913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,4096,0.008605333666006723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,6144,0.010033778018421596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,128,3072,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,3584,0.008332444561852349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,3584,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,3072,0.007717333734035492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,10240,0.012888000243239932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,128,2048,0.005283555636803309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,2560,0.006960888703664144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,2048,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,2048,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,128,1536,0.004587555511130227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,1536,0.006945778098371293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,128,1024,0.0038648889296584656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,128,2560,0.005746666755941179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,256,0.006086222413513396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,2560,0.007393777370452881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,128,768,0.003547555456558863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,1024,0.005621333503060871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,128,512,0.003244444520937072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,768,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,512,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,512,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,1536,0.005973333285914526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,256,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,256,0.0052915554907586836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,1024,0.006520000182920032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,128,0.004951111144489712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,128,64,0.0028826666788922418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,128,256,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,128,32,0.002867555452717675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,51200,0.03994222150908576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,64,65536,0.01447111037042406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,128,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,65536,0.054894222153557666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,64,0.004937777916590373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,51200,0.044052445226245456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,5120,0.008832000195980072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,64,12288,0.009737778041097853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,16384,0.01757955551147461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,64,10240,0.00903733323017756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,768,0.0052675555149714155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,12288,0.014484445254007975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,64,8192,0.008013333711359236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,64,7168,0.009022222624884712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,10240,0.01312444441848331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,64,6144,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,8192,0.011436444189813403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,64,16384,0.009367999931176504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,64,5120,0.008030222521887885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,64,4096,0.007325333025720384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,32,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,64,3584,0.006918222126033571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,5120,0.008797333472304875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,64,3072,0.006275555739800136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,4096,0.008062221937709385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,64,2560,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,64,51200,0.012462221913867526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,64,2048,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,3072,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,64,1536,0.004573333180612988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,2560,0.00700711127784517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,64,1024,0.0038888889054457345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,2048,0.0064693333374129395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,7168,0.01036800030204985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,64,768,0.003544888976547453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,3072,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,64,512,0.0032213332338465583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,6144,0.009958222508430481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,1024,0.005623999983072281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,768,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,512,0.005246222019195557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,64,64,0.002505777817633417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,256,0.005105777747101254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,64,32,0.0025457777082920074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,128,0.004941333499219683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,32,65536,0.015503999259736804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,64,0.00509333320789867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,32,0.004979555391603046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,32,51200,0.013447110851605734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,1536,0.006060444646411472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,64,256,0.0028764444092909494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,64,128,0.002847111059559716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,32,12288,0.009734222458468543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,16384,0.01755911111831665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,32,10240,0.009030222064918941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,12288,0.014506666196717156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,32,8192,0.008393777741326226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,10240,0.012861332959598966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,128,128,0.0026782221264309357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,65536,0.054479112227757774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,32,6144,0.008375999828179678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,7168,0.010618666807810465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,32,5120,0.007483555210961237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,32,4096,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,3584,0.007328888608349695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,6144,0.009880888793203566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,5120,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,32,3072,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,4096,0.007968000239796108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,32,2560,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,3584,0.007637333538797166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,32,7168,0.00904444439543618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,3072,0.007250666618347168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,32,16384,0.009353777600659264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,32,1536,0.004587555511130227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,2560,0.006648889018429651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,32,1024,0.0038746665749284956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,32,768,0.003536000019974179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,1536,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,1024,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,768,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,32,256,0.0028497777465316984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,512,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,32,2048,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,256,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,8192,0.011468444433477191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,51200,0.04415022333463033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,32,32,0.002546666603949335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,64,0.004937777916590373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,32,512,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,32,0.004972444640265571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,32,128,0.0028737777223189673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,32,64,0.00250133333934678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,2048,0.0063475556671619415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,225,32,3584,0.0069546666410234236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,51200,1.083595593770345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,16384,0.340831995010376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,16384,0.13992888397640652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,12288,0.30703732702467174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,16384,0.40437687767876523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,12288,0.2573404312133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,10240,0.2647164397769504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,12288,0.10611644718382095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,16384,0.31354043218824595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,10240,0.21132977803548178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,51200,1.2836266623602974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,12288,0.23915288183424208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,8192,0.21750844849480522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,8192,0.07329422235488892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,51200,0.43021244472927517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,8192,0.17973600493537056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,7168,0.19018222226036918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,8192,0.16171467304229736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,10240,0.08971999751196967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,7168,0.15789688958062068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,6144,0.16402488284640843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,6144,0.1325191126929389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,7168,0.1447191105948554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,128,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,5120,0.13815911610921225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,6144,0.0573271115620931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,6144,0.12523110707600912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,4096,0.11891288227505153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,5120,0.10900888840357463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,7168,0.0651457773314582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,5120,0.048651556173960366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,4096,0.09261244535446167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,3584,0.10010577572716607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,5120,0.10727377732594807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,4096,0.08476889133453369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,3584,0.03649955656793382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,10240,0.19815288649664986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,3584,0.07658666372299194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,3072,0.06831822130415174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,2560,0.07555555635028414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,3072,0.03267022305064731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,3072,0.06560977962281969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,2560,0.028684443897671167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,2560,0.059501330057779946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,51200,0.9745048946804471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,4096,0.04155911008516947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,2560,0.057095997863345675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,1536,0.04663733310169644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,2048,0.025273778372340735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,3072,0.09179200066460504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,2048,0.04773422082265218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,1536,0.039573334985309176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,1536,0.019825778073734708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,1536,0.03973244296179877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,1024,0.030768887864218816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,1024,0.015858666764365304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,768,0.0266951107316547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,2048,0.05024444394641452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,768,0.027228444814682007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,1024,0.0344142218430837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,768,0.014190221826235453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,512,0.021682666407691106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,512,0.023392889234754775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,512,0.013092444174819522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,256,0.012096000214417776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,768,0.025269332859251235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,3584,0.07977333333757189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,128,0.011658666862381829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,128,0.017885333961910672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,1024,0.032348444064458214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,64,0.009350222018029954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,128,0.011774222056070963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,32,0.010272888673676385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,128,0.01759999990463257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,256,0.018964444597562153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,64,0.01790488925245073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,512,0.019305777218606737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,256,0.012120889292822944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,2048,0.059911111990610756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,65536,1.1211973826090496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,51200,0.8809039857652453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,256,0.019546666079097323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,65536,1.0914195378621419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,32,0.017975111802419026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,65536,0.42696356773376465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,16384,0.2822008927663167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,51200,0.8353617986043295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,12288,0.21594844924079049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,16384,0.2713288995954725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,51200,0.33750579092237687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,12288,0.19916889402601454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,12288,0.08710844649208917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,16384,0.2549662325117323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,10240,0.20279022057851157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,10240,0.16910933123694527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,8192,0.15144977304670545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,10240,0.07119911246829562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,16384,0.113063997692532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,7168,0.1458293331993951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,8192,0.135897782113817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,8192,0.05850222375657824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,10240,0.1679848829905192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,6144,0.10928355322943793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,7168,0.12444978290134007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,8192,0.134624891810947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,12288,0.19160889254675972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,7168,0.11414044433169895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,6144,0.10315910975138347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,5120,0.09255022472805446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,6144,0.10208088821834987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,5120,0.038880000511805214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,4096,0.08058133390214708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,7168,0.052096890078650586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,4096,0.07192799780103895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,6144,0.04633333285649618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,5120,0.08717777993943955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,3584,0.06670133272806804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,4096,0.03360177742110358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,3072,0.058499554793039955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,3584,0.02998222245110406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,51200,0.7857795291476779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,65536,1.026400883992513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,3072,0.03272799981964959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,3072,0.05483377642101712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,3072,0.05571644173728096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,4096,0.07039644320805867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,2560,0.024060444699393377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,2048,0.04343555702103508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,2560,0.04825511243608263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,2048,0.020469332734743755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,3584,0.06314400169584486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,1536,0.03240977724393209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,2560,0.05253333184454176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,1536,0.032700445916917585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,1536,0.016754666964213055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,1024,0.025432000557581585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,1536,0.03456444541613261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,2560,0.05111111203829447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,768,0.01923199991385142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,1024,0.013996443814701505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,1024,0.02529688841766781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,1024,0.02812266680929396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,512,0.013186666700575085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,768,0.012895111408498553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,2048,0.041414221127827965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,768,0.02049511174360911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,256,0.009983110758993361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,512,0.011736888852384357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,512,0.020962667134073045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,256,0.016050666570663452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,128,0.009002666506502364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,256,0.01715022159947289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,3584,0.0648968882030911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,128,0.0151573336786694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,768,0.02442399991883172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,64,0.007407111426194509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,512,0.017690666847758822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,128,0.015445333388116626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,64,0.014975999792416891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,32,0.015224888920783997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,65536,0.38694310188293457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,128,0.010727110836240979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,65536,0.3638755480448405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,32,0.007660444411966536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,256,0.010684444672531553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,51200,0.310208002726237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,65536,0.14678666326734754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,16384,0.10968444744745891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,5120,0.08613333437177871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,51200,0.11678577793969049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,16384,0.09235822492175633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,12288,0.08394400278727214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,16384,0.04134133458137512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,16384,0.08771644698248969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,10240,0.06498399708006117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,12288,0.03257955445183648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,12288,0.06999555561277601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,12288,0.0686275561650594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,2048,0.04067911042107476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,10240,0.0281368891398112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,10240,0.05890578031539917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,10240,0.05815111266242134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,7168,0.05023911264207628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,8192,0.04927555720011393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,8192,0.04921333326233757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,7168,0.021648888786633808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,65536,0.3340897825029161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,7168,0.044850667317708336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,51200,0.2845395670996772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,8192,0.05684266487757365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,5120,0.037442667616738215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,6144,0.019129777948061626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,6144,0.0381368895371755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,6144,0.03984444340070089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,5120,0.01703377730316586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,5120,0.03269244564904107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,5120,0.03494577937655979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,7168,0.043235556946860425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,4096,0.015203555425008139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,3584,0.026966222458415564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,51200,0.2647413412729899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,3584,0.024691555235120986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,3072,0.0243493335114585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,3584,0.027535999814669292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,3584,0.013809777796268463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,8192,0.023950222465727065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,2560,0.020185778538386028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,3072,0.021736888421906367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,3072,0.02516355613867442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,4096,0.02777422302299076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,2048,0.016608888904253643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,4096,0.030127998855378892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,2560,0.012119111087587146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,2560,0.022663111488024395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,1536,0.01365244471364551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,2048,0.01112711098459032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,2048,0.01676266723208957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,2048,0.019881778293185763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,1536,0.014191110928853353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,4096,0.030087110069062974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,1536,0.01761511133776771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,1024,0.011737777955002256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,1024,0.008176888856622908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,6144,0.042796442906061806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,768,0.010258666343159145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,768,0.01310400002532535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,512,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,768,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,3072,0.013089777694808112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,512,0.009352888498041365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,256,0.005361777626805835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,512,0.0073315559162033936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,768,0.009384888741705153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,2560,0.01924888955222236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,128,0.004886222382386525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,256,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,256,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,256,0.010089777410030365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,64,0.004217777815130022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,128,0.00701422244310379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,1024,0.010763555765151978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,64,0.008222222328186035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,32,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,512,0.01142133358452055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,1024,0.01516266663869222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,65536,0.2929333315955268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,51200,0.2301431232028537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,65536,0.29128800498114693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,128,0.008190222084522247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,32,0.004547555413511065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,16384,0.08342222372690837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,51200,0.10206577512953018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,51200,0.22734043333265516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,16384,0.07682577768961589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,65536,0.31517688433329266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,65536,0.128565337922838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,16384,0.038600001070234514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,1536,0.009119999905427298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,12288,0.0304115547074212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,10240,0.0507351126935747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,10240,0.05003733436266581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,16384,0.08483378092447917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,12288,0.06424355506896973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,8192,0.044271111488342285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,10240,0.02585244509908888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,10240,0.05682133303748237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,12288,0.05851733022265964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,8192,0.04193244377772013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,12288,0.06769511434766981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,7168,0.019996444384257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,6144,0.033597333563698664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,7168,0.04321600000063578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,51200,0.2479271094004313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,128,0.010028444230556488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,6144,0.03280977739228143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,6144,0.018067555295096505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,6144,0.0388062232070499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,7168,0.03767199979888068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,8192,0.04776711265246073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,5120,0.028042667441897925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,5120,0.033795555432637535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,7168,0.03699288765589396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,3584,0.021332444416152105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,4096,0.02346311178472307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,4096,0.0296053323480818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,3584,0.013737777868906656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,5120,0.029085334804322984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,3072,0.018935999936527677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,8192,0.021928000781271193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,4096,0.02460977766248915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,3072,0.019066666563351948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,2560,0.015989333391189575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,3072,0.024703999360402424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,4096,0.014701333310869006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,2560,0.01702933344576094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,2048,0.014039999908871122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,2560,0.021851556168662176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,3584,0.021130666136741638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,3584,0.02672622270054287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,1536,0.01200888885392083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,2048,0.019518221418062847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,5120,0.01621866722901662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,1024,0.009649777577983009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,1536,0.00869333330127928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,1536,0.017265778448846605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,2560,0.011749332977665795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,768,0.008348444269763099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,1024,0.007689778175618913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,1024,0.014135999812020196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,768,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,2048,0.010197333163685268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,512,0.006311111152172089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,768,0.013000888956917657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,3072,0.012697777814335294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,256,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,512,0.008055110772450766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,512,0.01088711122671763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,1024,0.010411555568377176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,128,0.004575110971927643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,256,0.010102222363154093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,768,0.008758222063382467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,128,0.00756533361143536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,128,0.009720000127951304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,128,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,1536,0.012598222328556908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,64,0.007368889119890001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,32,0.007322666545708974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,65536,0.23321334520975748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,2048,0.014843554960356818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,65536,0.2474426693386502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,64,0.004233777936961916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,32,0.00424888895617591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,51200,0.18881332874298096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,65536,0.12467555205027263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,16384,0.07349244753519694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,51200,0.19347733921474883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,51200,0.09946844312879775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,16384,0.0654639999071757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,12288,0.05490755703714159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,16384,0.08453333377838135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,51200,0.2432462109459771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,12288,0.05038133263587952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,10240,0.04597422149446276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,12288,0.029079112741682265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,12288,0.06601155466503568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,10240,0.042913777960671314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,10240,0.024804444776640996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,8192,0.03736266824934218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,10240,0.05691288577185737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,8192,0.021376889612939622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,8192,0.036312000619040594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,8192,0.047692444589402944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,256,0.0069857777820693115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,7168,0.03205777870284186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,7168,0.019513777560657926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,16384,0.03687289026048448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,7168,0.042295111550225147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,512,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,6144,0.0280728903081682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,5120,0.024069334069887798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,6144,0.01759377784199185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,256,0.007600000335110559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,5120,0.024109333753585815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,5120,0.01588444411754608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,4096,0.01993600030740102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,5120,0.03323200013902452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,65536,0.3118355539109972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,3584,0.01742488808102078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,4096,0.01440533333354526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,4096,0.020555555820465088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,7168,0.03166933192147149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,3072,0.015498666299713982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,6144,0.03786755601565043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,3584,0.02643377747800615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,3072,0.01218933363755544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,2560,0.014092443717850579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,3072,0.024040000306235418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,6144,0.028335111008750066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,2560,0.01072266697883606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,2560,0.01519377695189582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,2560,0.021143999364640977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,4096,0.02847822176085578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,1536,0.010036444498433007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,3584,0.013343999783198038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,2048,0.009646221995353699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,2048,0.01923644377125634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,1024,0.00833155545923445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,1536,0.008371555142932469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,3072,0.017045333981513977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,1536,0.016513778103722464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,768,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,1024,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,1024,0.009389333426952362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,2048,0.01201600001917945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,1024,0.01349511080318027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,768,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,768,0.011751111182901593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,2048,0.013463111387358772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,3584,0.018780445059140522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,512,0.007094221810499827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,1536,0.011418666276666852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,256,0.007012444237867991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,128,0.004267555558019214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,256,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,256,0.009736888938479954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,512,0.005578666511509154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,128,0.006980444822046492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,128,0.009695111049546136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,32,0.0042168887125121225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,512,0.0075288886825243635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,64,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,32,0.006979555719428592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,65536,0.20123022132449678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,256,0.004572444491916233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,65536,0.0772542224989997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,65536,0.20283199681176078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,65536,0.17109244399600557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,51200,0.06211911307440864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,16384,0.054562668005625405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,512,0.011063999599880643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,16384,0.05441866649521721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,16384,0.02408711115519206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,16384,0.04827733172310723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,12288,0.04307911131117079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,51200,0.15670488940344915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,12288,0.04225511021084256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,10240,0.0346959994898902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,12288,0.03871555460823907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,51200,0.1382568942175971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,51200,0.16803733507792154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,10240,0.01720355616675483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,10240,0.03605866763326857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,10240,0.034068445364634194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,8192,0.01515111161602868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,8192,0.030393779277801514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,7168,0.026365333133273657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,12288,0.01958666741847992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,7168,0.0266942216290368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,6144,0.022966222630606756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,8192,0.03057066599527995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,6144,0.0236835562520557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,6144,0.013036444783210754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,5120,0.01962311069170634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,7168,0.014144889182514615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,5120,0.021765333082940843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,6144,0.024096000525686476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,5120,0.011758222348160215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,4096,0.016597333881590102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,4096,0.017815111411942374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,7168,0.026379555463790894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,4096,0.01902933253182305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,3584,0.014575999644067554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,4096,0.011047110789351992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,3584,0.01625155574745602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,768,0.00832799987660514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,3584,0.010055111514197456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,3072,0.013087111214796701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,3072,0.01482488877243466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,3072,0.016908443636364408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,2560,0.011399110986126794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,3072,0.009033777647548253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,2560,0.013395555317401886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,64,0.0038986665507157645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,2048,0.010217777556843227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,2560,0.008216000265545314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,5120,0.020609777834680345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,2048,0.014013333453072442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,1536,0.008443555898136562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,2048,0.007671111159854465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,1536,0.010047111246320937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,3584,0.018179555733998615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,1536,0.012507555385430654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,1536,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,1024,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,768,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,1024,0.010501333408885531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,768,0.007630222373538547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,768,0.009681777821646797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,512,0.004592888885074192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,2560,0.014922666880819531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,1024,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,256,0.00388355553150177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,512,0.006302222195598815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,512,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,512,0.009082666701740688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,1024,0.008186666501892937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,128,0.0039013334446483185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,256,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,128,0.006311111152172089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,64,0.003540444291300244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,768,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,8192,0.029102222787009344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,64,0.006663110935025745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,32,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,2048,0.011744889120260874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,65536,0.17990667290157744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,256,0.006275555739800136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,128,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,128,0.008649778034951951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,32,0.003555555724435382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,51200,0.15455022123124865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,51200,0.05917155742645264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,16384,0.052638222773869835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,256,0.00812355594502555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,51200,0.12479821840922038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,16384,0.04691466689109802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,65536,0.07368533478842841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,12288,0.04126222266091241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,12288,0.03894133369127909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,10240,0.034062223301993474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,16384,0.05297688974274529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,10240,0.035058667262395225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,10240,0.016680889659457736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,10240,0.03291822142071194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,8192,0.028557333681318495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,65536,0.1657573382059733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,12288,0.018908444378111098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,8192,0.014694222145610385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,8192,0.029179556502236262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,7168,0.026013332936498854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,7168,0.026554667287402686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,7168,0.013613333304723104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,6144,0.02164266672399309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,6144,0.023035556077957153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,8192,0.029224889145957098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,5120,0.01824711097611321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,12288,0.04045333464940389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,5120,0.019871999820073444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,5120,0.011632888681358762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,4096,0.01640533407529195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,5120,0.021393777595625982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,7168,0.02517333295610216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,4096,0.017264000243610807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,3584,0.014877332581414117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,51200,0.15665688779619005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,3584,0.015855110353893705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,3584,0.009373333719041612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,3584,0.017655111021465726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,3072,0.013047110703256396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,6144,0.02420088814364539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,6144,0.01276088919904497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,2560,0.011412444214026133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,3072,0.009191110730171204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,65536,0.19771644804212782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,2560,0.01310755560795466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,2560,0.007979555262459649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,2048,0.010028444230556488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,2560,0.014826666977670459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,2048,0.007637333538797166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,2048,0.0114720000161065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,2048,0.013374221821626028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,3072,0.01481599940194024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,4096,0.018956444329685636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,1536,0.009679111341635386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,16384,0.02275200022591485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,1024,0.006993778049945831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,1536,0.011848889291286469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,1024,0.006782222125265334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,768,0.005962666538026597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,1536,0.008599110775523717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,768,0.007336888876226213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,4096,0.010048888623714447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,512,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,768,0.00960622231165568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,1536,0.007184888753626082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,512,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,256,0.004247111164861255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,1024,0.010438222024175854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,256,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,256,0.00871377769443724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,256,0.006083555519580841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,768,0.006612444503439798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,128,0.006245333287451003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,64,0.003552888830502828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,128,0.008671110702885522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,32,0.003587555554178026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,64,0.00638133328821924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,32,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,65536,0.16554666890038386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,65536,0.1582568883895874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,512,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,51200,0.1315822203954061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,65536,0.07046755817201403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,65536,0.1517395575841268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,3072,0.01646311084429423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,128,0.003926222109132343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,16384,0.04586133360862732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,51200,0.05766133467356364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,51200,0.12150400214725071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,128,0.0063288890653186375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,12288,0.03742311067051358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,1024,0.008058666355080074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,16384,0.046909332275390625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,10240,0.031042665243148804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,12288,0.018167111608717177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,12288,0.0340577761332194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,12288,0.03748800026045905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,8192,0.02644088864326477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,10240,0.02899555696381463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,51200,0.1233182218339708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,10240,0.03252266512976752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,7168,0.023404444257418316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,8192,0.014310222533014087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,16384,0.02202400068442027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,512,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,6144,0.020357333951526217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,7168,0.013191111385822296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,7168,0.026086222794320848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,6144,0.019525332583321463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,5120,0.01711466742886437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,10240,0.01625155574745602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,5120,0.01720711092154185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,8192,0.024262222978803847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,5120,0.021128000484572515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,4096,0.014895111322402954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,5120,0.011004444625642566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,7168,0.021760000122918025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,3584,0.013436444103717804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,4096,0.015009777413474189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,4096,0.018650665879249573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,6144,0.012277333272827996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,6144,0.02353333267900679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,3584,0.013808888693650564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,3584,0.0176862229903539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,3072,0.012800888882742988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,2560,0.011023110813564725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,4096,0.009903999666372934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,2560,0.011415999796655444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,2560,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,3584,0.00903733323017756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,2560,0.01462399959564209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,3072,0.008711111214425828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,2048,0.009780444204807281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,3072,0.016173332929611206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,1536,0.008181333541870117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,2048,0.0074373334646224976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,1536,0.007177777588367462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,1024,0.006653333289755716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,1536,0.012049777640236748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,3072,0.012104888757069906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,1024,0.007322666545708974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,1024,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,16384,0.04349866509437561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,2048,0.01309333327743742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,768,0.006824888702895906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,512,0.004400889078776042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,768,0.006446222050322427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,1536,0.008674666285514832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,768,0.009703111317422655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,256,0.003871110992299186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,512,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,2048,0.0094035557574696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,512,0.008997333546479544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,768,0.0052871112194326185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,256,0.006113777971929974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,256,0.006000888844331105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,256,0.00850133349498113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,128,0.005984000033802456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,64,0.0031902222997612427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,512,0.006642666541867786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,64,0.006120000034570694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,1024,0.010250666903124915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,128,0.0035822221802340615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,32,0.006218666831652324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,8192,0.028143111202451918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,128,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,128,0.008064889245563084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,65536,0.06788622008429633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,51200,0.12632266680399576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,32,0.003220444545149803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,51200,0.05525955557823181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,51200,0.12040978007846409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,51200,0.12050755818684895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,65536,0.15730044576856825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,12288,0.033913777934180364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,65536,0.15266756216684976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,12288,0.032776888873842024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,10240,0.028714666763941448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,12288,0.017589333984586928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,65536,0.15148711204528809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,16384,0.04232711262173123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,8192,0.023450666003757056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,10240,0.015797333584891427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,16384,0.042302221059799194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,16384,0.021213332811991375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,8192,0.013855111267831592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,8192,0.02779200010829502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,7168,0.021407999926143225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,7168,0.0251555558707979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,12288,0.03687289026048448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,6144,0.01809866726398468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,10240,0.027787556250890095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,10240,0.032857778999540545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,6144,0.019189332922299702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,16384,0.04570133487383524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,7168,0.020950223008791607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,5120,0.010670222342014313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,4096,0.014152000347773233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,5120,0.020695999264717102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,7168,0.012894222305880653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,4096,0.014825777875052558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,3584,0.01310755560795466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,4096,0.009341333475377824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,6144,0.011854222251309289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,5120,0.015581332974963717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,8192,0.02383022175894843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,3072,0.011083555718262991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,3584,0.013733333183659447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,3584,0.017314665847354464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,5120,0.016894222961531747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,3072,0.008100444244013892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,3072,0.012409778104888068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,3072,0.016217778126398723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,2048,0.009118222528033787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,2560,0.007939555578761632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,2560,0.011104889214038849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,2560,0.014132445057233175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,4096,0.018182223041852314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,2048,0.007369777394665613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,2048,0.00979466653532452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,2048,0.012763555679056378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,1024,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,1536,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,1536,0.011477332976129321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,2560,0.01037599974208408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,1024,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,1024,0.006675555474228329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,1024,0.009815110928482479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,768,0.0064444442590077715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,1536,0.007979555262459649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,6144,0.023157333334287006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,1536,0.006989333364698622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,512,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,256,0.003950222084919612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,512,0.009023110899660323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,256,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,768,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,256,0.008391111261314815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,768,0.009687110781669617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,3584,0.008663111262851292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,128,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,64,0.003539555602603488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,128,0.005986666513813867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,32,0.003540444291300244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,768,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,65536,0.11794043911827935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,64,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,32,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,256,0.00599644457300504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,128,0.003607999947335985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,51200,0.0890551143222385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,65536,0.11592622598012288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,65536,0.14414221710628933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,128,0.008360889222886827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,51200,0.09003111388948228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,16384,0.03687466515435113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,51200,0.11443911658393012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,16384,0.020576889316240948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,12288,0.030113776524861652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,16384,0.04359733396106296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,65536,0.0656319989098443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,12288,0.025286222497622173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,12288,0.016915554801623028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,12288,0.03436888919936286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,51200,0.0524088872803582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,10240,0.021918222308158875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,8192,0.021439111895031396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,10240,0.030339555607901678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,512,0.004588444613748127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,8192,0.01890222231547038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,7168,0.018553778529167175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,8192,0.013615999784734515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,8192,0.02635555631584591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,10240,0.0255377772781584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,512,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,7168,0.0124177775449223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,7168,0.023706666297382776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,6144,0.01555999947918786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,6144,0.011189333266682096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,5120,0.014663999279340109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,6144,0.021799999806616042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,4096,0.01275199982855055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,5120,0.010042666561073726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,5120,0.013793778088357715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,5120,0.01956533392270406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,4096,0.012238221863905588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,4096,0.008801778157552084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,6144,0.016553777787420485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,4096,0.017646221650971305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,3072,0.008959111240175035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,3584,0.011056889262464313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,3584,0.016249777542220224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,10240,0.015205333630243937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,3072,0.010485333700974783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,3072,0.015375110838148328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,2560,0.0092275556590822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,3584,0.00961599995692571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,2048,0.007458666960398356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,2560,0.013117333253224691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,7168,0.017072889539930556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,3584,0.008379555410808986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,1536,0.006269333263238271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,2048,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,2048,0.012096000214417776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,2560,0.008025777836640676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,1024,0.00490044429898262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,1536,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,1536,0.011100444528791638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,1024,0.006612444503439798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,768,0.004269333349333869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,2560,0.007686222592989604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,2048,0.007335999773608313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,768,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,512,0.0041440000964535605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,768,0.009033777647548253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,3072,0.008153777983453538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,1536,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,256,0.0035386664999855887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,512,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,512,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,512,0.008750222623348236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,128,0.0032355555643637977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,1024,0.009720000127951304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,256,0.00794311116139094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,128,0.005633777628342311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,128,0.005987555616431766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,128,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,768,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,32,0.0031999999450312722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,65536,0.10619999965031941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,64,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,32,0.005647999958859549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,1024,0.006677333265542984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,256,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,65536,0.06467555628882514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,65536,0.14425510830349392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,64,0.0032417778339650896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,51200,0.09068799681133694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,16384,0.033611555894215904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,51200,0.11444532871246338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,16384,0.02019733356104957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,12288,0.028840889533360798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,256,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,65536,0.11339733335706924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,51200,0.08004888561036852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,10240,0.02458133300145467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,12288,0.016744888491100736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,12288,0.034338666333092585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,16384,0.0317831105656094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,51200,0.0513555539978875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,8192,0.020264888803164165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,10240,0.015146666102939181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,10240,0.030176000462638006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,16384,0.03164711263444688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,8192,0.013110222087966071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,8192,0.026133333643277485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,12288,0.025105777713987563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,7168,0.02342133389578925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,6144,0.016018667154841952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,10240,0.021964444054497614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,6144,0.01549600064754486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,7168,0.01817244456874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,6144,0.02183466653029124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,6144,0.010991999672518836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,7168,0.016926222377353244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,5120,0.01367022262679206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,4096,0.011101333631409539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,5120,0.019113777412308585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,4096,0.011777777638700275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,4096,0.01718844473361969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,8192,0.018681777848137748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,4096,0.008817777865462834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,5120,0.014129777749379476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,3072,0.009871111147933537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,3584,0.008378666308191087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,3584,0.010709332923094431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,3584,0.016201777590645682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,2560,0.00889155517021815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,3072,0.009690666364298927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,3072,0.01480977733929952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,2560,0.007713778151406183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,2048,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,2048,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,7168,0.01180622229973475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,2048,0.012033777932325998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,2048,0.007350222104125553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,1536,0.006611555400821898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,2560,0.0090524446633127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,2560,0.012791111237472959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,1024,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,1536,0.007017778025733099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,1536,0.01071911139620675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,1024,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,768,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,1024,0.009756444229020013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,5120,0.009488888912730748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,768,0.006303999986913469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,512,0.004222222086456087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,768,0.006243555496136348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,1536,0.007303111255168915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,512,0.006270222365856171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,256,0.0038720000949170855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,3072,0.008010666403505538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,256,0.005576000031497743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,256,0.005982222242487802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,16384,0.04301066531075371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,128,0.0035342222286595237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,512,0.005965333431959152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,512,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,64,0.0032195556494924757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,64,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,32,0.00319377767542998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,256,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,3584,0.010767999622556897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,65536,0.0820559991730584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,32,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,128,0.005468444277842839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,65536,0.10979021920098199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,51200,0.07070755296283297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,128,0.005968889014588461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,65536,0.1433617803785536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,16384,0.03234577841228909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,51200,0.050564444727367826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,51200,0.08572977781295776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,51200,0.11404355367024739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,768,0.00906311141120063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,12288,0.02629866699377696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,16384,0.04338399900330437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,12288,0.024637333220905725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,1024,0.006607111129495833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,10240,0.020929776959949072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,65536,0.06323822339375813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,10240,0.014847111370828418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,10240,0.021454221672481958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,8192,0.016301333904266357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,10240,0.02976800004641215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,16384,0.0200284438000785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,8192,0.012818666795889536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,8192,0.018398221996095445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,8192,0.02602311140961117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,12288,0.016536888149049547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,7168,0.011579555769761404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,12288,0.034183998902638756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,7168,0.02345955537425147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,6144,0.010375111467308467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,6144,0.015283554792404175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,6144,0.02160888910293579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,16384,0.03030222323205736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,5120,0.013331555657916598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,4096,0.010282666318946415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,5120,0.018958222534921434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,7168,0.016792888442675274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,6144,0.01423911088042789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,4096,0.011433777709801992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,4096,0.008760000268618265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,5120,0.013108444710572561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,4096,0.017062221964200337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,3584,0.01038933296998342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,7168,0.014523555835088095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,3584,0.015184889237085978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,5120,0.00943288869327969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,3072,0.009406222237481011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,3072,0.008019555773999956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,2560,0.00831911133395301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,3072,0.014144000079896716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,3584,0.009740444521109263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,2048,0.007267555428875818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,2560,0.00868800034125646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,2560,0.012652444342772165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,3584,0.008400888906584846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,3072,0.008660444782839881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,2048,0.011418666276666852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,128,0.00798044436507755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,1536,0.010759111079904767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,1536,0.00700711127784517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,1024,0.004765333400832282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,2560,0.007675555845101674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,768,0.004222222086456087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,1024,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,1024,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,1024,0.009332444104883406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,2048,0.008014221986134848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,512,0.0038844446341196695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,768,0.00902133352226681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,1536,0.007340444458855524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,512,0.005982222242487802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,256,0.0034462221794658234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,512,0.008362666600280339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,256,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,256,0.005616888817813661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,256,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,2048,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,128,0.0053288886944452924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,64,0.0034053333931499054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,128,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,32,0.0032231110251612137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,512,0.006059555543793573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,64,0.005590222362014983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,65536,0.07970222499635485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,32,0.005618666609128316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,128,0.0032231110251612137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,51200,0.06150933106740316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,65536,0.10614133543438381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,1536,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,65536,0.14355022377438015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,16384,0.02606666584809621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,51200,0.08348533180024888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,51200,0.11415821976131862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,16384,0.030010667112138536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,16384,0.01962222158908844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,16384,0.042372445265452065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,65536,0.06193333201938205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,12288,0.024059555596775476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,10240,0.019582221905390423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,768,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,128,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,51200,0.04943022131919861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,8192,0.01535377734237247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,10240,0.020854221449957952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,10240,0.02977422210905287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,12288,0.02179733415444692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,7168,0.013495999905798169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,8192,0.01237511138121287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,8192,0.025897777742809717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,7168,0.011400000088744693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,6144,0.012430222498046027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,10240,0.014146667387750415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,6144,0.014783110883500842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,6144,0.010136889086829292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,5120,0.01090844472249349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,7168,0.016249777542220224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,5120,0.012695111334323883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,5120,0.009391110804345872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,4096,0.010545777777830759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,8192,0.017940445078743827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,4096,0.011023110813564725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,6144,0.021352888809310064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,3584,0.00977422214216656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,4096,0.008716444174448649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,4096,0.016237333416938782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,3584,0.010216889282067617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,3072,0.009346666435400644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,3584,0.008392888638708327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,5120,0.018240888913472492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,3072,0.008039111064540016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,2560,0.008325333396593729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,2560,0.00868444475862715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,768,0.006301333506902059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,2560,0.0076657773719893555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,2048,0.007544000115659501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,2560,0.012778666284349231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,3072,0.009388444324334463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,3072,0.013976888524161445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,2048,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,1536,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,2048,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,2048,0.011738667057620155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,7168,0.023376888699001733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,1024,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,1536,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,1536,0.010823110739390055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,1024,0.006614222294754452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,768,0.004594666676388847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,1024,0.009376000199053023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,768,0.006055999961164262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,768,0.009025777379671732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,512,0.003933333274390963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,768,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,1536,0.007232888705200619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,3584,0.015660444895426433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,512,0.006061333335108227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,512,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,1024,0.006450666735569636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,256,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,256,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,128,0.003185777821474605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,256,0.007975111405054728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,64,0.0028764444092909494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,128,0.005965333431959152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,32,0.0028844444702068963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,128,0.0053342220683892565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,128,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,65536,0.06287022431691487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,256,0.0035866668654812705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,512,0.008775999976529015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,65536,0.066520889600118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,51200,0.05080088973045349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,65536,0.14332444137997097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,51200,0.05277688966857063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,16384,0.022671111755900916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,51200,0.11373599370320638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,64,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,16384,0.020228443874253165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,16384,0.01942399972014957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,12288,0.020290666156344943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,65536,0.06087733639611138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,12288,0.01665511065059238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,12288,0.015535111228624979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,12288,0.0337066650390625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,51200,0.048046221335728966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,12288,0.033823112646738686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,10240,0.01347555551264021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,8192,0.014259555273585849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,10240,0.02938400043381585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,32,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,7168,0.012872888810104795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,8192,0.011693333586057028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,8192,0.025311110748185053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,7168,0.01179377817445331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,10240,0.0192284451590644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,10240,0.014853333433469137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,6144,0.01070666644308302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,5120,0.010618666807810465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,6144,0.010119111173682744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,8192,0.013047999805874295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,5120,0.009407111340098912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,4096,0.00907199995385276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,5120,0.017993777990341187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,6144,0.011781333221329583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,4096,0.00832444429397583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,3584,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,4096,0.01622577839427524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,6144,0.0211173329088423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,16384,0.0426879988776313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,3584,0.008000888758235509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,3072,0.007679111427730984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,3584,0.008356444537639618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,5120,0.009294222626421187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,2560,0.0069422221018208405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,3072,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,3072,0.007333333293596904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,7168,0.010899555351999072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,3072,0.014345778359307183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,2048,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,2560,0.007661333514584436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,4096,0.008719111482302347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,1536,0.005246222019195557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,2560,0.01277599980433782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,2048,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,2048,0.006670222100284364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,1024,0.004345777961942884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,1536,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,1536,0.006265777680608962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,1536,0.010383110907342698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,3584,0.015042665931913586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,768,0.004229333251714706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,1024,0.005922666854328579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,1024,0.009692444569534725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,7168,0.023033777872721355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,768,0.0063475556671619415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,768,0.005655111124118169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,768,0.009015111459626092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,512,0.0053013331360287136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,256,0.0032355555643637977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,256,0.0052471111218134565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,512,0.003839111162556542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,256,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,128,0.0031902222997612427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,512,0.005967999911970562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,2048,0.011374221907721626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,128,0.004938666605287128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,128,0.005980444451173146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,128,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,1024,0.006599999964237213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,65536,0.052741332186592944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,64,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,32,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,2560,0.006993778049945831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,51200,0.041189332803090416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,65536,0.0588888857099745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,512,0.008431110945012834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,51200,0.050478223297331065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,51200,0.04741511080000135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,16384,0.018564444449212816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,51200,0.11344799730512832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,16384,0.018566222654448617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,16384,0.019746666153271992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,12288,0.016207110550668504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,16384,0.042169776227739125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,256,0.008047111332416534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,10240,0.01510933372709486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,12288,0.015375110838148328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,12288,0.016230222251680162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,65536,0.14270488421122232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,12288,0.03329244587156508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,8192,0.014833778142929077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,10240,0.01314933349688848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,64,0.0031991110493739447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,7168,0.013510222236315409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,8192,0.01142133358452055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,8192,0.025090666280852422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,7168,0.01145155562294854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,7168,0.010764444039927589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,7168,0.0221404449807273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,65536,0.06422577963935004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,6144,0.010040000081062317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,5120,0.010016000105275048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,10240,0.01457244488928053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,10240,0.028951999213960435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,8192,0.01242844429281023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,5120,0.00924711094962226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,5120,0.009339555270142024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,5120,0.018280888597170513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,6144,0.011424000064531961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,4096,0.00868266655339135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,4096,0.01573600040541755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,6144,0.010259555445777046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,3584,0.008033778104517195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,6144,0.020390222469965618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,3072,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,3584,0.014831999937693277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,4096,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,3072,0.00739911115831799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,3072,0.013794666363133324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,4096,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,2560,0.0069982219073507525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,2048,0.006602666858169768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,2560,0.007675555845101674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,2560,0.012763555679056378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,2048,0.007297777467303806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,2048,0.006656888872385025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,2048,0.011408888631396823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,3072,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,2560,0.007285333342022366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,1024,0.004575110971927643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,1536,0.0069884442620807225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,1536,0.01071733319097095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,1024,0.005899555567238066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,768,0.00423555572827657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,3584,0.008348444269763099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,1536,0.005622222191757626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,768,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,512,0.003940444439649582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,768,0.006333333336644703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,768,0.00868088917599784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,512,0.0052675555149714155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,512,0.006252444452709622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,256,0.003532444437344869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,512,0.008410666551854875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,1024,0.006610666712125142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,1024,0.009367999931176504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,256,0.006011555592219035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,256,0.004948444250557157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,256,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,32,0.0032311110860771606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,128,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,128,0.004955555415815777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,32,0.0028853333658642242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,128,0.008018666671382057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,1536,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,32,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,3584,0.007984889050324758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,65536,0.061531556977166064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,128,0.0028560000161329904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,51200,0.03076710965898302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,65536,0.14253955417209202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,64,0.002936888900068071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,16384,0.01573422220018175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,65536,0.03753955496682061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,51200,0.11290666792127822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,64,0.004936000125275718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,12288,0.013456000222100152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,16384,0.019261333677503798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,16384,0.04169777697987027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,65536,0.05776266919242012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,12288,0.016002666619088914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,10240,0.013245333400037555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,12288,0.014160000615649752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,12288,0.033223999871148005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,8192,0.012078222301271228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,10240,0.012862222062216865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,51200,0.048433777358796865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,10240,0.02883466747072008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,8192,0.011412444214026133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,8192,0.024381332927280005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,16384,0.016905777984195285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,7168,0.011058666639857821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,6144,0.011378666592968835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,7168,0.02221155497762892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,6144,0.010034666293197209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,10240,0.013791999883121915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,5120,0.010027555955780877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,6144,0.019708444674809773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,6144,0.010058666268984476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,7168,0.011350221931934357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,4096,0.008630221916569604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,5120,0.009356444080670675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,5120,0.009015999734401703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,3584,0.008319999608728621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,4096,0.008292444050312042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,4096,0.015592000550693937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,3584,0.007680888805124495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,3072,0.00737955586777793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,51200,0.04626399940914578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,8192,0.012086222569147745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,3072,0.013791110780504016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,2560,0.006606222026877933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,5120,0.017551110850440133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,7168,0.010768888725174798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,2560,0.007686222592989604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,2560,0.006959999601046245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,3584,0.008375110725561777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,2560,0.012102222277058495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,3584,0.014835556348164877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,1536,0.005014222115278244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,2048,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,2048,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,2048,0.011446221835083432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,3072,0.007328888608349695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,1024,0.0042133331298828125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,3072,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,4096,0.008692444198661381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,1024,0.005646222167544895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,1024,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,1024,0.009385777844323052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,2048,0.006136888845099344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,512,0.0035742223262786865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,768,0.006331555545330048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,768,0.005619555711746216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,768,0.00904444439543618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,256,0.0032115555885765287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,512,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,512,0.0052844443255000645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,512,0.00833777752187517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,128,0.0032302221904198327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,256,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,256,0.005250666704442766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,256,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,64,0.0028568889117903183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,128,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,128,0.005005333158704969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,32,0.003031999907559819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,128,0.008011555506123437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,768,65536,0.032596445745891996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,64,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,32,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,65536,0.05994400050905016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,1536,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,1536,0.010409778190983666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,51200,0.04738577869203356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,768,16384,0.01705777810679542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,51200,0.0458808872434828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,51200,0.11242577764723037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,16384,0.018608000543382432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,768,12288,0.014533332652515836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,16384,0.041703111595577665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,65536,0.057486222849951855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,1536,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,12288,0.014146667387750415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,12288,0.03231733375125461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,768,0.0038933331767717996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,10240,0.013451555536852943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,16384,0.016896888613700867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,10240,0.028006222512986925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,768,8192,0.012244444754388599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,10240,0.012854222622182635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,768,51200,0.02674399978584713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,768,7168,0.011431111229790581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,8192,0.011392888923486074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,8192,0.011741333537631564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,8192,0.02455111179086897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,768,6144,0.010381333529949188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,7168,0.0107342220014996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,7168,0.02164711058139801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,6144,0.00996888925631841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,768,5120,0.009703111317422655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,6144,0.01999288962946998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,5120,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,12288,0.014845333165592618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,7168,0.01092177795039283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,4096,0.00814488861295912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,6144,0.010035555395815108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,768,3584,0.007269333634111617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,65536,0.14274222320980495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,768,10240,0.014299554957283868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,3584,0.008302222523424361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,768,3072,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,3584,0.014552889598740472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,4096,0.00871288859181934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,4096,0.015843555331230164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,3072,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,768,2560,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,3072,0.007967111137178209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,2560,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,2560,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,768,4096,0.007676444119877285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,2560,0.012425777812798819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,2048,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,2048,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,768,1536,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,2048,0.01145155562294854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,1536,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,1536,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,3072,0.013813333378897773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,3584,0.007746666669845581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,768,2048,0.0052888890107472735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,768,768,0.003910222401221593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,1024,0.009407999614874521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,768,0.005648889061477449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,768,0.008671110702885522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,768,512,0.003555555724435382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,768,1024,0.004273777620659934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,1536,0.010457778142558204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,512,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,1024,0.0058417779703934984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,512,0.008372444245550368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,1024,0.0064080001579390625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,256,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,768,128,0.0032524443748924467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,256,0.007985778152942657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,768,0.006361777583758037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,128,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,768,64,0.002887111157178879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,128,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,768,32,0.002842666581273079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,512,0.006073777874310811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,5120,0.009342222577995723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,512,65536,0.025090666280852422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,64,0.004945777770545748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,32,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,256,0.004993777722120285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,65536,0.05699200100368924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,65536,0.14241777526007757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,51200,0.04683644572893778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,51200,0.04494933287302653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,512,16384,0.011918221910794577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,51200,0.11255555682712132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,5120,0.017549332645204332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,512,12288,0.013502221968438892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,16384,0.04145333170890808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,512,51200,0.019814221395386588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,65536,0.05838221973843045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,12288,0.014840000205569796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,512,10240,0.012418666647540199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,12288,0.014142221874660917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,12288,0.032553778754340276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,512,8192,0.010392888552612728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,10240,0.0129439996348487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,10240,0.013264888690577613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,10240,0.02789688772625393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,512,7168,0.010428444378905825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,8192,0.011425777441925473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,8192,0.01162755572133594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,16384,0.01813155578242408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,8192,0.02402577797571818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,512,6144,0.009748444788985783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,7168,0.01072266697883606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,7168,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,512,5120,0.009356444080670675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,6144,0.010048888623714447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,6144,0.00976533359951443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,6144,0.020008888509538438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,5120,0.009031111167536842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,128,0.005975111077229182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,5120,0.017539555827776592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,768,256,0.0031884445084465873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,4096,0.00812355594502555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,4096,0.00869333330127928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,512,3584,0.0070168889231152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,4096,0.015511110424995422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,7168,0.02169599963559045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,512,3072,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,3584,0.007660444411966536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,3584,0.014832889040311178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,5120,0.009379555781682333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,3072,0.007292444507280986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,3072,0.008020444048775567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,3072,0.013800000150998434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,16384,0.016882666283183627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,512,2048,0.005764444255166584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,2560,0.012466666599114736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,2048,0.007355555891990662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,3584,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,512,1536,0.004945777770545748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,2048,0.011416888899273343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,1536,0.006928000185224745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,1536,0.006072000082996156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,512,1024,0.0041786668201287585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,1536,0.010407110883129967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,1024,0.006657777975002925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,1024,0.005699555493063397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,2560,0.006996444529957241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,1024,0.009351111120647853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,768,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,768,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,768,0.008772444393899705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,512,4096,0.00796977761718962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,512,0.0053022222386466126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,512,256,0.0033031110134389666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,512,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,512,768,0.003932444585694207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,256,0.005032888717121548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,512,128,0.0029235554652081597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,256,0.008032889001899296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,2048,0.0063440000845326324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,512,512,0.003547555456558863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,512,64,0.002867555452717675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,128,0.005910222315125995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,512,32,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,128,0.007954667011896769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,512,2560,0.005982222242487802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,64,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,512,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,32,0.004941333499219683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,256,51200,0.019499555230140686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,65536,0.051446222596698336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,256,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,51200,0.045853333340750806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,51200,0.04123555620511373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,128,0.00499555551343494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,256,65536,0.020593777298927307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,16384,0.01759111053413815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,256,12288,0.00923111124171151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,16384,0.016869333055284288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,2560,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,65536,0.05721244547102186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,256,10240,0.009391999906963771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,12288,0.014176888598336114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,12288,0.03194311261177063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,256,16384,0.011628444823953839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,51200,0.1124231153064304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,256,8192,0.00868622213602066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,10240,0.027748444014125403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,8192,0.011413333316644033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,16384,0.04060533311631944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,8192,0.0236791107389662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,12288,0.014525334040323893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,7168,0.010739555789364709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,256,6144,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,7168,0.01071911139620675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,10240,0.013162666724787818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,10240,0.012892444100644855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,256,5120,0.0083137775460879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,6144,0.009752000371615091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,6144,0.01998044384850396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,256,7168,0.008217777642938826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,256,4096,0.007595555649863348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,5120,0.00941422250535753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,5120,0.017553778158293832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,65536,0.14219466845194498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,4096,0.008676444490750631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,4096,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,4096,0.015554666519165039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,3584,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,6144,0.010040000081062317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,256,3072,0.0063795554969045855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,8192,0.011434666812419891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,5120,0.00889244427283605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,256,2560,0.00592533333433999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,3072,0.007363555332024892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,3072,0.013812444276279874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,256,3584,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,7168,0.021644444929228887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,2560,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,256,2048,0.005199111170238919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,2560,0.007383999725182851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,3584,0.0083137775460879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,2048,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,2048,0.011063999599880643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,1536,0.005976888868543837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,1536,0.010392000277837118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,256,1024,0.003896000070704354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,2560,0.012427555190192329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,1024,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,256,1536,0.004606222112973531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,256,768,0.0035662220584021676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,1024,0.00657066661450598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,1024,0.009424000150627559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,3072,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,256,512,0.0031946665710873077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,1536,0.006640000061856375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,512,0.005285333428117964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,256,256,0.0029102222373088202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,512,0.006092444476154115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,2048,0.006681777950790193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,256,0.005231110999981563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,256,128,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,768,0.005649777750174205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,768,0.008561777571837107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,128,0.005024888863166173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,256,64,0.002512000087234709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,128,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,256,32,0.002576888849337896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,512,0.008368888662921058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,3584,0.01482488877243466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,128,65536,0.01609866652223799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,32,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,256,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,256,0.007796444826655918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,65536,0.04931911163859897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,768,0.006275555739800136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,65536,0.14191910955641004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,128,16384,0.009713778065310584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,51200,0.03983911209636264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,128,0.008013333711359236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,51200,0.11186044745975071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,128,12288,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,16384,0.016731555263201397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,64,0.005149333427349727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,12288,0.014158222410413953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,65536,0.05643288956748115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,12288,0.03204888767666287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,51200,0.04461333486768934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,10240,0.013177778157922955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,128,8192,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,10240,0.012803555362754397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,16384,0.017252445220947266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,16384,0.04035022192531162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,128,7168,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,8192,0.011432888607184092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,8192,0.02403822210099962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,128,10240,0.009118222528033787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,7168,0.010752889017264048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,7168,0.010420444111029306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,7168,0.02164799968401591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,6144,0.009879999690585667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,128,5120,0.007984889050324758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,6144,0.01976799964904785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,8192,0.011377777490350934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,5120,0.008919999831252629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,128,51200,0.014212444424629211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,5120,0.017543110582563613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,128,6144,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,4096,0.008047999607192146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,4096,0.008676444490750631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,128,3584,0.007033777733643849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,6144,0.009678222239017487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,10240,0.027522666586769953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,128,3072,0.006993778049945831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,3584,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,3584,0.014800000521871777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,12288,0.014519110321998596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,3072,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,128,2560,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,3072,0.013481777575280933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,2560,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,2560,0.006999111009968652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,128,2048,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,2560,0.012105777859687805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,2048,0.00703999979628457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,3584,0.008336000144481659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,128,1536,0.004549333204825719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,2048,0.011422221859296164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,128,4096,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,1536,0.0069511110583941145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,128,1024,0.0038942222793896994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,1536,0.010077333284748925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,1024,0.006330666856633292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,1024,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,128,768,0.003539555602603488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,4096,0.015503111812803479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,2048,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,768,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,128,512,0.00319377767542998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,3072,0.007356444166766272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,512,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,1024,0.009363555245929295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,512,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,768,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,256,0.004957333207130432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,768,0.006324444380071428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,128,128,0.0029235554652081597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,1536,0.005978666659858491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,256,0.007675555845101674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,128,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,128,64,0.0025173332542181015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,512,0.008383110993438298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,128,32,0.0028968888024489083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,64,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,64,65536,0.014484445254007975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,32,0.004965333475006951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,64,51200,0.012811555630630918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,64,16384,0.009703111317422655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,65536,0.05447644326421949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,64,12288,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,128,0.005036444299750858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,51200,0.04398577743106418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,64,10240,0.007691555553012424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,64,8192,0.007382222347789341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,16384,0.01757244434621599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,64,7168,0.0069751110341813825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,10240,0.01314222233162986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,64,6144,0.006676444576846228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,7168,0.010484444598356882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,256,0.005922666854328579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,6144,0.009736888938479954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,64,4096,0.006649777707126405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,5120,0.00868088917599784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,64,3584,0.006324444380071428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,128,256,0.0031386667655573953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,4096,0.008035555481910706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,12288,0.014465777410401238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,64,2560,0.006319111006127463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,3584,0.007646222081449296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,64,2048,0.006053333481152852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,3072,0.0073315559162033936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,64,1536,0.004610666798220741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,2048,0.006640888750553131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,64,1024,0.00387999994887246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,1536,0.006021333237489064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,128,0.007677333222495184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,64,3072,0.005980444451173146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,1024,0.005619555711746216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,64,512,0.0034764446318149567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,768,0.005452444569932089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,8192,0.01145511120557785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,64,128,0.00253866674999396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,512,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,64,5120,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,256,0.004975999808973736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,64,32,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,128,0.005070222334729301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,32,65536,0.014345778359307183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,64,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,64,768,0.003559111307064692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,32,51200,0.012409778104888068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,32,0.004969777746333016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,32,16384,0.009011555876996782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,32,12288,0.007976888782448238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,65536,0.05486577749252319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,64,256,0.0029848888516426086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,51200,0.043896890348858304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,16384,0.01752800080511305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,12288,0.014470222923490735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,32,7168,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,10240,0.01311466677321328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,32,6144,0.00664444433318244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,7168,0.010435555544164447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,32,5120,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,6144,0.009871111147933537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,32,4096,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,32,10240,0.007363555332024892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,5120,0.008676444490750631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,32,3584,0.006292444550328785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,32,3072,0.006312888943486744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,2560,0.006664888726340399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,32,2560,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,3584,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,8192,0.011420444481902652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,32,1536,0.004567999806669024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,2560,0.006675555474228329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,32,1024,0.0037431112594074676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,2048,0.006302222195598815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,32,768,0.003555555724435382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,1024,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,32,8192,0.007347555624114142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,768,0.005395555661784277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,4096,0.008032889001899296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,32,256,0.002862222285734283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,64,64,0.0025128889828920364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,32,128,0.0025253333151340485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,32,2048,0.006153777655627992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,512,0.0052782222628593445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,32,64,0.002692444456948174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,3072,0.007356444166766272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,32,32,0.002518222149875429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,128,0.004958222309748332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,64,0.004954666727119022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,1536,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,224,32,512,0.0031893334041039148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,16384,0.40014489491780597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,256,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,51200,1.2448488871256511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,51200,1.0758577982584636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,16384,0.14232444763183594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,12288,0.3009342352549235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,51200,0.42289333873324925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,12288,0.10443644391165839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,12288,0.25291911760965985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,10240,0.21095644103156197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,8192,0.20525777339935303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,10240,0.08933066659503514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,7168,0.17662844392988417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,8192,0.07250044743220012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,8192,0.17206666204664442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,6144,0.15129066838158503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,7168,0.15399021572536892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,10240,0.2732124328613281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,6144,0.05687644746568468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,6144,0.1300897730721368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,4096,0.10688800281948513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,5120,0.048713776800367564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,5120,0.10756888654496934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,3584,0.1021066639158461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,4096,0.08943821986516316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,3072,0.08133066362804837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,3584,0.03644444545110067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,32,0.004981333182917701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,3584,0.0773128867149353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,7168,0.06375910838445027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,2560,0.06708888875113593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,3072,0.032427555984920926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,2048,0.05584177705976698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,2560,0.028060442871517603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,2560,0.058648890919155545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,1536,0.04535466763708326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,2048,0.024653333756658766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,2048,0.04959555466969808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,1024,0.03231822119818793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,1536,0.019368888603316415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,4096,0.040806220637427434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,16384,0.34146489037407773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,1536,0.03914577762285868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,1024,0.015662221444977652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,512,0.018416888184017606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,768,0.01418133411142561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,256,0.013212444053755866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,768,0.024689777029885188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,512,0.024899555577172175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,3072,0.06743199957741632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,256,0.018917333748605516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,128,0.011757333245542316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,128,0.018232888645595975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,64,0.00831911133395301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,768,0.02555111050605774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,1024,0.030397332376903955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,64,0.01789422167672051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,512,0.01309511148267322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,32,0.01790577835506863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,5120,0.12824177742004395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,32,0.008676444490750631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,128,0.010079999764760336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,65536,1.2805795669555664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,51200,1.006813367207845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,65536,1.0565510855780709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,51200,0.8369555473327637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,51200,0.3265555434756809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,12288,0.24822666909959581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,12288,0.08323911163542005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,10240,0.1996142201953464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,16384,0.27868355645073783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,65536,0.42375911606682676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,8192,0.16379733880360922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,16384,0.32542578379313153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,10240,0.16822400357988146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,10240,0.07049689027998182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,7168,0.14353333579169378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,8192,0.05888533592224121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,8192,0.13501600424448648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,7168,0.05139288968510098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,6144,0.12349955240885417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,7168,0.11844089296129014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,5120,0.10650222169028388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,16384,0.10945155885484482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,6144,0.04537333382500542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,4096,0.08725688854853313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,5120,0.03842666745185852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,5120,0.0854862199889289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,4096,0.07108977768156263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,3584,0.029746668206320867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,3072,0.06915644142362806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,256,0.012101333174440594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,12288,0.19877333111233184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,6144,0.10690044694476658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,3072,0.05492977632416619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,3584,0.07599377632141113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,4096,0.03311911225318909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,2048,0.04722933305634392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,2560,0.04759022262361315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,3584,0.06249777475992838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,1536,0.032285332679748535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,3072,0.026639110512203638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,1024,0.026728888352711994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,2560,0.023982221881548565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,1024,0.014182221558358936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,768,0.020689777202076383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,768,0.012864888542228274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,512,0.015241778559154935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,2560,0.05570311016506619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,512,0.011781333221329583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,512,0.021006221572558086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,256,0.01110577748881446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,2048,0.020106666617923312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,128,0.009003555609120263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,768,0.020412445068359375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,128,0.015094222293959724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,64,0.007295110987292395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,1536,0.016548444827397663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,32,0.00867733359336853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,64,0.014984000060293408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,2048,0.040264887942208186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,256,0.016015110744370353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,32,0.014881778094503613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,65536,0.381951994366116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,51200,0.3019173410203722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,128,0.01068800025516086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,65536,0.14256355497572157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,16384,0.10285333130094741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,51200,0.27429599232143825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,51200,0.11275555027855767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,16384,0.04042933384577433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,1024,0.02474044428931342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,10240,0.0629591080877516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,12288,0.03223911258909438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,12288,0.06914133495754667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,8192,0.054930667082468666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,10240,0.02802844511138068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,10240,0.058225777414109975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,65536,0.34798399607340497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,7168,0.04657511247528923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,8192,0.023600889576805964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,6144,0.041328890456093684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,7168,0.02142488956451416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,7168,0.042612443367640175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,256,0.010553778045707278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,6144,0.019161777363883126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,5120,0.01689511040846507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,5120,0.032489776611328125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,4096,0.027448889282014634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,12288,0.07616977559195624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,3584,0.02611644400490655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,1536,0.03714666763941447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,6144,0.03780000077353583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,4096,0.015190222197108798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,3072,0.02329866588115692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,3584,0.014027555783589682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,2560,0.01939733326435089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,4096,0.029932445949978296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,3072,0.02181600034236908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,2048,0.01610844499535031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,2560,0.01930488811598884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,8192,0.04904355605443319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,1536,0.01310755560795466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,2048,0.010996444357766045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,2048,0.0166311115026474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,1024,0.010467555787828235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,1536,0.009073778159088558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,1536,0.014101333088344999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,1024,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,1024,0.011779555843936073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,3072,0.01312888910373052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,512,0.006991110742092133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,768,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,2560,0.012092444631788464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,256,0.005094222310516569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,512,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,512,0.010503999888896942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,128,0.004608889006906086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,256,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,256,0.008325333396593729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,64,0.0038986665507157645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,128,0.007024889190991719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,32,0.004221333397759331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,768,0.008960888617568547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,3584,0.024319110645188227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,32,0.008040889269775814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,65536,0.2868737909528944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,5120,0.03462844424777561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,51200,0.22729865709940592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,128,0.008073777788215214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,64,0.008055110772450766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,16384,0.07656177547242907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,65536,0.30084533161587185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,51200,0.2231848769717746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,12288,0.06018310785293579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,16384,0.03799110982153151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,16384,0.07583644655015734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,16384,0.09068533447053696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,12288,0.03016977839999729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,10240,0.025803556044896443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,10240,0.04978311061859131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,8192,0.0437466667758094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,768,0.010037333601050908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,8192,0.021887999441888597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,7168,0.03891288903024461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,8192,0.04158400164710151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,6144,0.03299199872546726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,7168,0.01995111174053616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,65536,0.12716710567474365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,6144,0.018163555198245578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,6144,0.03230844603644477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,5120,0.01626133256488376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,4096,0.0243848893377516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,51200,0.10134400261773004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,3584,0.021184888150956895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,4096,0.014522666732470194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,4096,0.023383110761642456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,3072,0.018931556079122756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,3584,0.013444444371594323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,3584,0.020979555116759405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,2560,0.016103999482260812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,3072,0.012678222523795234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,3072,0.018961777289708454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,2048,0.013806222213639153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,2560,0.011746666497654386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,5120,0.02777155571513706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,1536,0.01167377746767468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,2048,0.01462933255566491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,1536,0.008551999926567078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,1024,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,5120,0.028449777099821303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,1024,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,768,0.008460444708665213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,2560,0.016920889417330425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,2048,0.010048888623714447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,512,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,768,0.007384888827800751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,768,0.008959111240175035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,256,0.005084444251325396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,512,0.007339555356237624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,7168,0.03665688965055678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,128,0.004226666771703296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,256,0.007486222518814935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,1024,0.010381333529949188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,128,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,128,0.007297777467303806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,32,0.003925333420435588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,64,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,32,0.007408888803588019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,1536,0.012582222620646158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,10240,0.05011999938223097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,12288,0.05769689215554131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,51200,0.20004355907440186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,65536,0.12443733215332031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,65536,0.24389778243170845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,16384,0.07900710900624593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,51200,0.1901306708653768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,51200,0.09843555423948501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,16384,0.03659644391801622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,16384,0.06515733400980632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,10240,0.050344000260035195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,12288,0.028565333949195013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,12288,0.049513777097066246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,8192,0.041132443481021456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,10240,0.04264800084961785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,8192,0.021350221501456365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,64,0.003927999900446998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,12288,0.05834133095211453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,7168,0.036079999473359846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,6144,0.03261155552334256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,7168,0.01960088809331258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,65536,0.26422754923502606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,5120,0.028789334826999243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,6144,0.027748444014125403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,5120,0.016178665889634024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,8192,0.035783999496036105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,4096,0.0236151119073232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,4096,0.014192889134089151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,3584,0.02095288866096073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,512,0.009125333693292405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,10240,0.024713777833514746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,3584,0.018593778212865193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,3072,0.018269333574506972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,3072,0.012136889000733694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,2560,0.016192888220151264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,3072,0.016858667135238647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,2560,0.010728889041476779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,2048,0.014491554763582019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,256,0.007000889215204451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,3584,0.013165333204799227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,2048,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,2048,0.01332266628742218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,5120,0.023670223024156358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,1536,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,4096,0.020392000675201416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,6144,0.017543110582563613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,1024,0.009016888837019602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,768,0.007636444436179266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,1024,0.00776977754301495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,1536,0.01184266640080346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,512,0.005973333285914526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,2560,0.0151591118839052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,512,0.008369777765538957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,256,0.004600000050332812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,256,0.0069875551594628235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,128,0.004278222305907143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,768,0.007329777710967594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,768,0.008025777836640676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,64,0.00425155543618732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,128,0.006730666591061487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,7168,0.031670222679773964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,32,0.004230222354332606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,512,0.00702311098575592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,65536,0.2300648954179552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,65536,0.19744533962673613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,65536,0.0761680006980896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,51200,0.184371550877889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,128,0.006937777830494775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,1024,0.009382222261693742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,16384,0.062126219272613525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,51200,0.062281780772738986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,12288,0.0487040016386244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,32,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,16384,0.05390577846103244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,10240,0.03817777832349142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,12288,0.019542222221692402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,12288,0.04137066668934292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,256,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,8192,0.03386488887998793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,10240,0.017208000024159748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,10240,0.035641776190863714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,8192,0.015154666370815702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,1536,0.01149511088927587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,8192,0.029776000314288672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,51200,0.15439644124772814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,7168,0.013935999737845527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,64,0.00693866651919153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,5120,0.021964444054497614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,16384,0.02422133253680335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,4096,0.018185777796639335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,5120,0.011756444142924415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,5120,0.020205333828926086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,4096,0.010813333094120026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,3584,0.0162764439980189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,4096,0.01758666667673323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,3072,0.01459200017982059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,3584,0.009759999811649323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,3584,0.016103999482260812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,2560,0.012510221865442065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,6144,0.02546311087078518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,3072,0.014828445182906257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,7168,0.026191999514897663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,6144,0.01296266665061315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,2560,0.013268444273206921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,1536,0.009117333425415887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,2048,0.007866666548781926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,2048,0.011800000237094032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,1024,0.0076248885856734375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,1536,0.007366221812036302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,1536,0.009689778089523315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,7168,0.02878755662176344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,1024,0.00832622249921163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,3072,0.009032888544930352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,512,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,768,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,768,0.0075342224703894714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,256,0.003926222109132343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,512,0.0063368889192740125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,512,0.007325333025720384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,128,0.0035582222044467926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,256,0.006234666539563074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,256,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,128,0.006255110932721033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,128,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,32,0.0035386664999855887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,64,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,2048,0.010819555156760745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,32,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,2560,0.008036444584528605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,6144,0.023217777411142986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,51200,0.1667297813627455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,65536,0.07251022259394328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,65536,0.1947253280215793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,16384,0.05617066886689928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,51200,0.15194222662183973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,51200,0.058485335773891874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,768,0.006083555519580841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,12288,0.04375822345415751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,16384,0.022760889596409265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,16384,0.052737779087490506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,12288,0.018616888258192275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,12288,0.040511998865339495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,8192,0.03144977821244134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,65536,0.20644266075558135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,8192,0.014807111687130399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,7168,0.02747822304566701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,64,0.0032071111102898917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,1024,0.006980444822046492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,7168,0.013791110780504016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,10240,0.03689510954750909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,6144,0.01259200026591619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,10240,0.0166284441947937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,5120,0.02052088909678989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,10240,0.03474044468667772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,4096,0.01714933415253957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,5120,0.011397333608733283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,8192,0.028866665230857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,3584,0.015178667174445258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,4096,0.010263111028406356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,3072,0.01389600005414751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,4096,0.017280000779363845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,3584,0.009369778136412302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,6144,0.023916444844669763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,2560,0.012085333466529846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,3584,0.015829333000712924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,2048,0.010431999961535135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,3072,0.014369777507252164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,2560,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,1536,0.008723555339707268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,2560,0.012903999951150684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,2048,0.0074471111098925276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,1024,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,1536,0.007143110864692264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,5120,0.0199635558658176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,1024,0.006762666834725275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,1024,0.008010666403505538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,768,0.006641777853171031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,512,0.00490399988161193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,6144,0.02267733381854163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,512,0.006993778049945831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,2048,0.01127822200457255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,512,0.0063288890653186375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,1536,0.009183111290136972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,256,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,768,0.005622222191757626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,256,0.006335999816656113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,64,0.003547555456558863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,128,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,32,0.003887999802827835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,3072,0.008432000047630733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,64,0.0063235556913746735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,32,0.006611555400821898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,256,0.004250666747490565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,51200,0.1454417837990655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,128,0.003903111235962974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,51200,0.05658666955100166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,128,0.006325333482689328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,7168,0.025602666868103877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,65536,0.17630578411950004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,16384,0.021856000026067097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,12288,0.040582223070992365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,16384,0.0432275566789839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,10240,0.035270220703548856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,12288,0.01812799937195248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,65536,0.06985955768161349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,8192,0.02908266584078471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,10240,0.016185777054892648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,10240,0.028332445356580947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,7168,0.025599999560250178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,8192,0.014243554737832812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,51200,0.12215110990736221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,768,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,6144,0.02402044501569536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,7168,0.013394667042626275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,7168,0.021552888883484736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,6144,0.012144889268610211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,6144,0.01924533314175076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,5120,0.01108000013563368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,5120,0.016968000266287062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,12288,0.033075554503334895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,3584,0.014742222097184924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,4096,0.009575111170609793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,4096,0.01483022173245748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,3072,0.01311911145846049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,3584,0.00904088881280687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,16384,0.05215200119548374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,3584,0.013831111292044321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,2560,0.011919111013412476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,3072,0.008442666795518663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,5120,0.019630221856964957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,3072,0.012725333372751871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,2560,0.007822222179836696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,1536,0.008996444443861643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,2048,0.007364444434642792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,2048,0.009621333744790819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,65536,0.1558364497290717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,1536,0.0069875551594628235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,1536,0.008343111309740279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,1024,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,768,0.0059262220230367445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,1024,0.007329777710967594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,8192,0.024093333217832778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,512,0.004573333180612988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,768,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,768,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,256,0.003879111260175705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,2560,0.011448000040319232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,512,0.006664000037643645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,128,0.003522666792074839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,256,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,1024,0.0069226668112807805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,64,0.0031999999450312722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,128,0.00599644457300504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,32,0.003536888708670934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,64,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,32,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,2048,0.010693333215183683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,512,0.0063164445261160536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,4096,0.016182222300105624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,51200,0.13055288791656494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,256,0.005973333285914526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,16384,0.04871288935343424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,65536,0.15102577209472656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,51200,0.1176275544696384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,51200,0.0546248886320326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,16384,0.020853334003024634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,16384,0.041758222712410815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,12288,0.01757066614098019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,12288,0.032002667586008705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,65536,0.15679644213782415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,8192,0.026664000418451097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,10240,0.02772533396879832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,7168,0.025143999192449782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,8192,0.013833777772055732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,8192,0.023623110519515142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,7168,0.013120888835854001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,7168,0.021013332737816706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,128,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,5120,0.018925334016482036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,6144,0.011728888584507836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,6144,0.01869333287080129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,4096,0.016730666160583496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,5120,0.010382222632567087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,5120,0.016562667157914903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,3584,0.015142222245534262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,4096,0.009454222189055549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,4096,0.014578666951921252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,3072,0.011721777419249216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,3584,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,3584,0.013479111095269522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,2560,0.010027555955780877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,12288,0.03723377651638455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,10240,0.03233244352870517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,2560,0.00775111135509279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,2560,0.010543111297819348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,10240,0.0158595558669832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,1536,0.007738666401969061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,6144,0.021164443757798936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,1536,0.008247111406591203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,1024,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,3072,0.012087110843923358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,2048,0.009206222163306342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,768,0.00518488883972168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,1024,0.00756888919406467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,2048,0.007325333025720384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,512,0.004551110996140374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,768,0.006304889089531369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,768,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,256,0.00398577791121271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,512,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,65536,0.0677057769563463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,3072,0.00813688917292489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,256,0.005992888990375731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,1536,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,64,0.0032400000426504347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,1024,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,32,0.003216000066863166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,64,0.005964444329341252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,2048,0.009696889254781935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,32,0.006234666539563074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,512,0.006600889066855113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,128,0.0035866668654812705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,65536,0.11409244272443984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,51200,0.10429155826568604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,256,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,16384,0.03971022367477417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,128,0.006002666635645761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,51200,0.052299555804994374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,16384,0.020277332928445604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,16384,0.03165955675972833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,10240,0.025184889634450276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,12288,0.02476444509294298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,65536,0.12505688932206896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,8192,0.021067554752031963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,10240,0.01531022290388743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,10240,0.02164977788925171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,7168,0.018564444449212816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,8192,0.013396444420019785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,8192,0.018639999959203932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,128,0.005964444329341252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,7168,0.016976000534163583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,51200,0.0880017810397678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,12288,0.016858667135238647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,5120,0.014672888649834527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,6144,0.015220445063379077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,5120,0.009694221946928237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,4096,0.015203555425008139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,65536,0.06536177794138591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,4096,0.008849778109126622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,3584,0.013229333692126803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,7168,0.012248888611793518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,6144,0.016363556186358135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,3072,0.012097777591811286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,3584,0.00850666645500395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,6144,0.011222222612963783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,3584,0.010785777535703448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,3072,0.007985778152942657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,3072,0.009767110976907942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,2048,0.010051555931568146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,2560,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,2560,0.008896888958083259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,1536,0.008325333396593729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,2048,0.007289778027269576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,2048,0.008011555506123437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,1024,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,1536,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,1024,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,2560,0.011380444798204633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,768,0.006335999816656113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,512,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,12288,0.030071109533309937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,512,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,256,0.0035946667194366455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,1024,0.006657777975002925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,256,0.005980444451173146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,128,0.0035146665241983202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,5120,0.013622221847375235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,128,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,128,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,512,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,32,0.00318755561278926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,64,0.005573333137565189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,768,0.005623111294375525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,32,0.005818666683302985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,256,0.005616888817813661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,768,0.006265777680608962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,65536,0.06388533115386963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,65536,0.11212710539499919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,16384,0.036954667833116316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,64,0.003198222153716617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,51200,0.08666577604081895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,16384,0.020087111327383254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,16384,0.030941334035661485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,65536,0.1137839953104655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,12288,0.024476443727811176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,51200,0.09238133165571426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,10240,0.021478222476111516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,8192,0.019848888119061787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,51200,0.05087911089261373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,12288,0.029859556092156306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,7168,0.017902221944597032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,12288,0.016688888271649677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,10240,0.024069334069887798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,6144,0.015850666496488783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,7168,0.011772444678677453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,6144,0.010424888796276517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,5120,0.013965333501497904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,10240,0.01518311103185018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,8192,0.012857777376969656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,4096,0.010801778071456485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,7168,0.01681688924630483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,4096,0.00870488915178511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,3584,0.011559111376603445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,4096,0.011845333708657159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,3584,0.00852177788813909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,5120,0.009411555197503831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,3072,0.010557333628336588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,5120,0.013456888496875763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,2560,0.00830844458606508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,3072,0.008021333151393467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,3072,0.009728888670603434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,2048,0.0074417781498697065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,2560,0.008795555267069075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,2048,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,1536,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,3584,0.010508444574144153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,1536,0.0070257774657673305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,1024,0.0058311112225055695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,4096,0.011759999725553723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,1024,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,768,0.004912000149488449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,6144,0.015232889188660515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,768,0.006317333214812809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,512,0.0041928887367248535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,768,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,8192,0.018453333112928603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,256,0.0035502223504914176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,512,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,512,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,2560,0.007654222349325816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,256,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,256,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,1024,0.006611555400821898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,128,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,64,0.002895111011134254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,128,0.0052897776994440295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,1536,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,32,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,65536,0.08999466896057129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,128,0.003212444484233856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,65536,0.10593066612879436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,51200,0.07146310806274414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,2048,0.007982222570313348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,32,0.0032275555034478507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,16384,0.02857511242230733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,51200,0.08334222104814316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,12288,0.02386133372783661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,64,0.005628444254398346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,10240,0.022712000542216833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,12288,0.016507556041081745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,12288,0.0239955551094479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,8192,0.0184897780418396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,10240,0.014825777875052558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,65536,0.06303022305170695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,10240,0.020928889513015747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,7168,0.014408000641398959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,8192,0.012476444244384766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,8192,0.018173333671357896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,1536,0.0073315559162033936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,51200,0.04981777734226651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,5120,0.014013333453072442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,6144,0.01480977733929952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,16384,0.020000888241661918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,4096,0.012107555237081317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,5120,0.009473777479595607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,5120,0.013111111190583972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,3584,0.01143555591503779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,4096,0.008721777962313758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,4096,0.011094222466150919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,3584,0.008345777789751688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,6144,0.012858666479587555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,3584,0.010083555347389644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,7168,0.011072888970375061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,2560,0.009159999589125315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,7168,0.016287111573749118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,2048,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,16384,0.030219554901123047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,2560,0.008800889054934183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,2048,0.008016889293988546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,1536,0.006180444525347815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,2048,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,1024,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,1536,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,1024,0.006635555376609166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,1024,0.00666222224632899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,768,0.004582222137186262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,768,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,2560,0.007687110867765214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,512,0.00423911131090588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,768,0.005980444451173146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,256,0.0035777779089079965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,512,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,512,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,1536,0.006960888703664144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,256,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,256,0.005619555711746216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,64,0.0032097777972618737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,128,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,3072,0.008033778104517195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,128,0.0052675555149714155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,65536,0.08254577716191609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,64,0.005631999837027655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,32,0.005592888842026393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,51200,0.0654017792807685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,65536,0.10360711150699192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,3072,0.010064889159467485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,51200,0.04878933231035868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,51200,0.0817448894182841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,32,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,6144,0.010336000058386061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,12288,0.022335110439194575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,16384,0.029635555214352075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,10240,0.01959733333852556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,12288,0.02336799932850732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,65536,0.061646223068237305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,8192,0.019357333580652874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,128,0.003212444484233856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,16384,0.027564444475703772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,7168,0.013750221994188098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,3072,0.009376889301670922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,7168,0.010877333581447601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,7168,0.016033778587977093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,6144,0.012441777520709567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,6144,0.010043555663691627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,10240,0.01426755554146237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,6144,0.014463999205165438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,5120,0.00942488925324546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,8192,0.017616000440385606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,16384,0.019641778535313077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,12288,0.01621955633163452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,4096,0.01072355525361167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,3584,0.00961511085430781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,5120,0.01092355532778634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,3584,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,3584,0.010051555931568146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,3072,0.008961777720186446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,8192,0.011750222080283694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,2560,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,3072,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,3072,0.009305777649084726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,2048,0.007318221860461765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,2560,0.00868800034125646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,4096,0.008671999805503422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,1536,0.006268444574541516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,2048,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,1536,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,1024,0.005125333451562458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,1536,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,5120,0.012103111379676394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,768,0.004570666700601578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,1024,0.006637333167923822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,1024,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,2560,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,512,0.00387999994887246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,768,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,256,0.003571555432346132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,512,0.006044444524579578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,2048,0.0073315559162033936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,10240,0.020624000165197585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,128,0.0032711111836963226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,64,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,128,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,32,0.0032213332338465583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,128,0.005325333525737126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,64,0.005427555491526921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,32,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,768,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,51200,0.055528889099756874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,65536,0.064374221695794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,512,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,16384,0.023744000328911677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,51200,0.04731111062897576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,51200,0.050817777713139854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,12288,0.018908444378111098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,16384,0.01949155496226417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,16384,0.019870221614837646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,65536,0.06525599956512451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,12288,0.015274667077594332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,12288,0.016166221764352586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,8192,0.014481777946154276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,10240,0.013162666724787818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,10240,0.014504889647165934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,7168,0.01347644461525811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,8192,0.0127688886390792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,256,0.005979555762476391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,6144,0.01238933371173011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,256,0.005576888720194499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,7168,0.01072533345884747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,7168,0.01166044423977534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,5120,0.012081777883900536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,6144,0.010072888599501716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,4096,0.00923644420173433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,5120,0.009372444616423713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,5120,0.009387555221716562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,3584,0.008375999828179678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,4096,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,65536,0.060319999853769936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,8192,0.011408000356621213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,3584,0.008346666892369589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,3584,0.008023111356629265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,2560,0.006952889263629913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,3072,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,3072,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,4096,0.010283555421564314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,2048,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,2560,0.007648888561460707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,10240,0.018926221463415358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,1536,0.0052471111218134565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,2560,0.0070364442136552595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,1024,0.004259555704063839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,2048,0.006683555742104848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,1536,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,768,0.003934222377008862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,1024,0.006614222294754452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,3072,0.007632888853549957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,1024,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,512,0.00388711111413108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,768,0.0063271112740039825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,768,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,256,0.0035946667194366455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,512,0.0053102220926019884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,256,0.005984000033802456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,128,0.0032311110860771606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,256,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,2048,0.007326222128338284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,128,0.005981333139869902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,32,0.003185777821474605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,128,0.004946666873163647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,1536,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,64,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,65536,0.05070844292640686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,32,0.005292444593376583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,51200,0.03999555442068312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,65536,0.06169244315889147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,6144,0.010037333601050908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,51200,0.048785779211256236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,4096,0.008678221868144141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,16384,0.018561777141359117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,16384,0.019260444574885897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,12288,0.016885333591037326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,12288,0.014888889259762235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,10240,0.015587555037604438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,51200,0.04666755596796671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,64,0.0028524444335036804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,10240,0.014157333307796054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,8192,0.013791110780504016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,10240,0.012854222622182635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,7168,0.012695999609099494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,8192,0.011424889167149862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,8192,0.011758222348160215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,6144,0.011054221954610614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,7168,0.01105955574247572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,65536,0.05827822287877401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,5120,0.010015111002657149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,6144,0.010035555395815108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,5120,0.009355555805895064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,5120,0.00924444446961085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,4096,0.008783111141787635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,3584,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,4096,0.008656889200210571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,4096,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,3072,0.007294221884674496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,7168,0.010763555765151978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,3584,0.008009778128729926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,512,0.005976000179847081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,2560,0.007247111035717859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,12288,0.015862221519152325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,3072,0.007305777735180325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,2560,0.007689778175618913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,2560,0.00701333334048589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,1536,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,2048,0.006613333192136552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,1536,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,1024,0.004574222283230888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,3584,0.008347555167145198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,1024,0.006609777609507243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,1024,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,2048,0.006335111127959357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,16384,0.018269333574506972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,768,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,1536,0.006220444622966979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,512,0.006263999889294307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,256,0.003560888684458203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,768,0.004583111239804162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,6144,0.009951111343171861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,128,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,256,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,512,0.00388711111413108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,64,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,128,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,128,0.004938666605287128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,32,0.0028568889117903183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,65536,0.03632444474432204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,512,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,64,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,32,0.005229333208666907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,65536,0.057461334599388965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,65536,0.05925866630342272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,16384,0.015495111544926962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,51200,0.04621422290802002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,768,0.006317333214812809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,2048,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,16384,0.018897778458065458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,12288,0.014150222142537435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,12288,0.015203555425008139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,51200,0.029776000314288672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,3072,0.008035555481910706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,8192,0.012307555311255984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,10240,0.013131555583741931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,51200,0.0469084448284573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,7168,0.011876444849703046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,12288,0.013109332985348172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,8192,0.011752888560295105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,6144,0.010395555860466426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,7168,0.01075466639465756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,7168,0.010750222537252637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,5120,0.011078221930397881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,10240,0.014127999544143677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,6144,0.010237333675225576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,6144,0.009717333647939894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,5120,0.009369778136412302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,5120,0.00904888908068339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,10240,0.012818666795889536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,3584,0.00832088871134652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,4096,0.008046222229798635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,256,0.005054222212897407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,3072,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,3584,0.008287111090289222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,3584,0.007689778175618913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,2560,0.0064035554726918536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,3072,0.007972444097201029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,3072,0.007328000333574083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,2048,0.005584888988071018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,2560,0.007680888805124495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,4096,0.008364444805516137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,1536,0.00563822231358952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,2048,0.007297777467303806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,2048,0.006618666566080517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,1024,0.004217777815130022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,1536,0.006980444822046492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,1536,0.006311999840868845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,768,0.003936888857020272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,16384,0.016890666551060148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,1024,0.005912000106440649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,512,0.0035386664999855887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,768,0.0063271112740039825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,768,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,256,0.003232888877391815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,512,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,2560,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,128,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,256,0.005924444645643234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,256,0.005299555758635203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,64,0.0028488888508743713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,128,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,32,0.0028515555378463534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,64,0.005081777771313985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,768,65536,0.02941688895225525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,32,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,512,0.005461333526505365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,65536,0.05714222457673815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,768,51200,0.02482755482196808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,65536,0.0579377810160319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,768,16384,0.016924444172117446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,4096,0.008676444490750631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,51200,0.045956446064843066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,8192,0.011462222370836469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,1024,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,16384,0.017909333109855652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,768,10240,0.01313066648112403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,12288,0.014191110928853353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,12288,0.014581332604090372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,768,8192,0.011720889144473605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,10240,0.012807110945383707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,768,7168,0.011354666617181567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,8192,0.01143022212717268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,8192,0.011406222151385413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,768,6144,0.010042666561073726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,7168,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,128,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,6144,0.010031999813185798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,6144,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,5120,0.009367999931176504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,768,4096,0.008773333496517604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,10240,0.013066666821638743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,4096,0.00868800034125646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,7168,0.010762666662534079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,768,3584,0.007162666983074612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,768,12288,0.014581332604090372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,51200,0.04587466518084208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,3584,0.007835555407736037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,3072,0.008055999875068665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,5120,0.009012444151772393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,4096,0.008045333127180735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,2560,0.007678222325113084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,3584,0.008303999900817871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,768,3072,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,2048,0.007296000089910295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,768,1536,0.004896000027656555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,3072,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,1536,0.006955555743641323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,768,1024,0.004214222232500712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,2560,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,1024,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,768,768,0.0038720000949170855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,16384,0.016917333006858826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,768,0.006528888725572162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,2048,0.006598222172922558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,768,2560,0.00591466658645206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,512,0.005987555616431766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,512,0.005295111073387994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,768,2048,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,768,5120,0.00889955543809467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,768,128,0.0031866667171319327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,256,0.005964444329341252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,768,512,0.0035582222044467926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,128,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,768,64,0.002945777856641345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,128,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,1536,0.006089777582221561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,64,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,512,65536,0.025625778569115534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,32,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,512,51200,0.02079288827048408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,1024,0.005668444352017508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,65536,0.05647199683719211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,512,16384,0.012132444315486483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,51200,0.04416888952255249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,51200,0.04518577787611219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,768,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,512,12288,0.010032888915803697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,16384,0.01684533390733931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,16384,0.01779911087618934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,512,10240,0.012115555504957834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,12288,0.01425155500570933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,12288,0.014504889647165934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,512,8192,0.010843555960390302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,10240,0.013092444174819522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,65536,0.05693333016501533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,512,7168,0.010391111175219217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,8192,0.011403555671374003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,8192,0.011369778050316704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,512,6144,0.009328000247478485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,7168,0.01071733319097095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,7168,0.010673777924643623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,768,32,0.0029279999434947968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,512,5120,0.009002666506502364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,6144,0.009714666340086196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,768,256,0.003240888938307762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,5120,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,5120,0.008943110704421997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,10240,0.013097777962684631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,512,3584,0.007244444555706448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,4096,0.008661333057615492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,512,3072,0.006607111129495833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,3584,0.008383110993438298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,3584,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,512,2560,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,3072,0.007289778027269576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,6144,0.010045333041085137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,2560,0.007615110940403408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,2560,0.0069413334131240845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,512,4096,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,2048,0.0074933336840735535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,2048,0.00630044440428416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,512,1024,0.0038844446341196695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,3072,0.008051555189821456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,1024,0.005707555760939916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,512,768,0.004223111189073986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,1024,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,512,2048,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,768,0.006321777900060018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,512,512,0.0032115555885765287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,768,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,512,256,0.0031991110493739447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,512,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,512,0.005344888816277186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,256,0.005990222096443176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,256,0.005060444275538127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,128,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,512,64,0.002847111059559716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,512,1536,0.0048942222363419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,512,32,0.0028924445311228433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,1536,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,512,128,0.0029102222373088202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,32,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,65536,0.049970666567484535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,128,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,65536,0.05582933293448555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,51200,0.040123555395338274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,256,16384,0.010433778166770935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,64,0.004939555707905027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,16384,0.016640888320075143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,256,12288,0.0101742222905159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,16384,0.01754488878779941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,12288,0.014193778236707052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,256,10240,0.009386666946940953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,256,51200,0.01969688965214623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,10240,0.012816888590653738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,256,8192,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,10240,0.013062222136391534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,8192,0.011407111254003314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,256,0.004982222285535601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,256,7168,0.009348444640636444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,256,65536,0.01993688941001892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,256,6144,0.009892444643709395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,7168,0.01074399964676963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,7168,0.010399111443095738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,12288,0.014151111245155334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,6144,0.009566222627957663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,5120,0.009360888765917884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,5120,0.008720888859695857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,1536,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,8192,0.01127644462717904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,4096,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,4096,0.008656889200210571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,3584,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,3584,0.007684444387753804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,51200,0.04458222124311659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,3072,0.007124444676770105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,256,2560,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,3072,0.007977777885066139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,4096,0.008073777788215214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,256,2048,0.005017777697907554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,2560,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,2560,0.006931555353932911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,256,1536,0.004567111117972268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,2048,0.007300444775157505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,256,5120,0.00868266655339135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,256,3072,0.0064071110553211635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,256,1024,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,1536,0.006803555621041193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,256,4096,0.007631111476156447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,1024,0.006353777729802662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,256,768,0.0035537779331207275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,1024,0.00572622236278322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,256,512,0.003206222214632564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,768,0.006317333214812809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,256,3584,0.006948444578382704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,768,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,256,256,0.0028942223224374983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,512,0.006121777825885349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,256,128,0.002847111059559716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,256,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,6144,0.010083555347389644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,256,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,128,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,128,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,64,0.004953777624501122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,128,65536,0.01682844426896837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,32,0.004960888789759742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,65536,0.04929244518280029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,128,51200,0.014394667413499622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,512,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,2048,0.006602666858169768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,256,64,0.002564444517095884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,128,16384,0.009720889230569204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,51200,0.042857779396904834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,128,12288,0.009695111049546136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,16384,0.01684711045689053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,16384,0.01723288827472263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,128,10240,0.009357333183288574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,12288,0.014152889450391134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,12288,0.014144889182514615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,128,8192,0.008349333372380998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,10240,0.012824888858530255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,8192,0.011367999845080905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,128,7168,0.00906488878859414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,8192,0.011074666347768573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,256,32,0.002526222210791376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,7168,0.01074577785200543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,7168,0.010375111467308467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,128,5120,0.008653333617581261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,6144,0.010028444230556488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,6144,0.009352888498041365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,128,4096,0.007626666790909237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,5120,0.009372444616423713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,5120,0.008652444514963362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,65536,0.05308444301287333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,128,3584,0.006934222247865465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,4096,0.007896888587209914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,128,3072,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,51200,0.03973333372010125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,3584,0.007326222128338284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,128,2560,0.005592888842026393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,3072,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,2560,0.007639110916190677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,2560,0.006646222124497096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,128,2048,0.004943999979231092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,10240,0.012827555338541666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,4096,0.00870222184393141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,128,1536,0.004588444613748127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,2048,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,128,1024,0.0038728887836138406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,1536,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,1536,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,1536,0.005998222364319696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,3072,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,1024,0.006672888994216919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,1024,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,128,512,0.0032346666687064697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,768,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,128,256,0.003208000005947219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,768,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,512,0.005212444398138259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,256,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,2048,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,128,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,128,0.004965333475006951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,128,6144,0.008671999805503422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,128,32,0.002826666666401757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,64,0.00499288903342353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,64,65536,0.024306666519906785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,512,0.006024000131421619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,64,51200,0.019609777463806998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,128,128,0.0028693332440323303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,64,16384,0.009343110852771336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,65536,0.05306222372584873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,64,12288,0.00942133367061615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,3584,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,51200,0.04251733422279358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,128,64,0.002536888958679305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,64,8192,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,128,768,0.0035546666218174826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,12288,0.014255111416180929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,10240,0.012816000315878125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,64,6144,0.008422222402360704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,32,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,7168,0.010393777655230628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,64,5120,0.007632888853549957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,256,0.004949333353175057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,64,4096,0.007626666790909237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,5120,0.008839111361238692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,64,3584,0.006888888776302338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,4096,0.008000888758235509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,64,7168,0.00902133352226681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,8192,0.011061333119869232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,64,2560,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,3072,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,64,2048,0.004954666727119022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,6144,0.009444444543785518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,2560,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,64,10240,0.008681777450773451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,64,1024,0.00387999994887246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,1536,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,64,3072,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,64,512,0.003208000005947219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,1024,0.005677333308590784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,64,256,0.0028799999919202593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,768,0.005547555370463266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,64,128,0.002847111059559716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,256,0.004931555440028508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,64,64,0.002528888897763358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,128,0.004908444566859139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,2048,0.006301333506902059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,16384,0.016869333055284288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,32,65536,0.02439022229777442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,64,0.00509688879052798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,32,51200,0.02200355629126231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,3584,0.007671111159854465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,32,16384,0.012808889150619507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,65536,0.05255200134383308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,512,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,51200,0.042466666963365346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,32,10240,0.009015999734401703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,64,1536,0.004263999975389904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,16384,0.01703466640578376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,32,8192,0.008319999608728621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,12288,0.014166222678290473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,10240,0.012786666552225748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,32,6144,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,64,768,0.003563555578390757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,32,5120,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,6144,0.009374221993817223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,32,4096,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,32,12288,0.00958577791849772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,32,3584,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,4096,0.008027555214034187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,32,7168,0.009038222332795462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,3584,0.007328888608349695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,32,2560,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,8192,0.011346666349305047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,32,2048,0.005010666532648934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,7168,0.010390222072601318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,32,1536,0.004228444563017951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,2560,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,32,1024,0.0038888889054457345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,2048,0.006409777949253718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,5120,0.008672888908121321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,1536,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,32,512,0.0032551110618644287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,1024,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,32,256,0.003010666618744532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,64,32,0.0025128889828920364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,32,128,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,512,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,32,64,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,256,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,32,0.004640000147951974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,32,32,0.0025333333760499954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,64,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,32,0.004750222381618288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,768,0.005417777846256892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,3072,0.007152888509962294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,128,0.004931555440028508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,51200,0.8662417729695638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,32,768,0.0035253332720862497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,16384,0.3957342306772868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,16384,0.2788044346703423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,193,32,3072,0.0063075555695427795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,12288,0.29846755663553876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,16384,0.1383271084891425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,51200,0.4132533338334825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,12288,0.2050337791442871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,10240,0.23936798837449816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,12288,0.10665688912073772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,10240,0.18013866742451987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,10240,0.0881999995973375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,8192,0.2010044389300876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,8192,0.14615821838378906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,8192,0.07340888844596015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,7168,0.17113244533538818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,10240,0.1965866618686252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,7168,0.12127821975284153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,8192,0.1572151051627265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,7168,0.06436444653405084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,6144,0.10485866996977065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,7168,0.1381431155734592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,51200,1.2332258224487305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,6144,0.05596711238225301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,6144,0.12097599771287705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,6144,0.15099288357628718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,5120,0.08840799993938869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,4096,0.10663200087017483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,16384,0.30653778711954754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,5120,0.048215111096700035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,4096,0.07293244202931722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,5120,0.12388177712758382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,4096,0.04065333472357856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,3584,0.09014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,51200,0.9704009162055122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,3584,0.036031110419167414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,3072,0.0840231113963657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,12288,0.24225955539279512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,3072,0.03237866693072849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,3072,0.06548444430033366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,2560,0.06881778107749091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,4096,0.08244088623258802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,2048,0.053117334842681885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,2560,0.04943111207750109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,2560,0.0566248893737793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,2048,0.041399998797310725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,2048,0.024586665961477492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,2048,0.047832889689339526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,1536,0.03295822276009454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,1536,0.03978577918476529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,1536,0.019242667489581637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,1024,0.030633777379989624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,3584,0.0751511123445299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,3072,0.056571556462181934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,768,0.023853333459960088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,1024,0.03167288833194309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,768,0.02090577781200409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,768,0.02739288906256358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,512,0.017264889346228707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,768,0.014189332723617554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,512,0.01738577749994066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,3584,0.06350044409434001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,512,0.02350844442844391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,1536,0.041197333070966936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,512,0.013217777841620974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,256,0.012088889049159156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,256,0.015678221980730694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,128,0.01036177741156684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,256,0.019256888164414298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,2560,0.028174221515655518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,64,0.009111111362775167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,128,0.014879110786649915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,128,0.017494221528371174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,32,0.009834667046864828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,5120,0.1024177802933587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,256,0.012015110916561551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,64,0.014492443866199918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,32,0.015807999504937064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,128,0.011744889120260874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,65536,1.0928924348619249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,65536,0.4136977725558811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,51200,0.8518071174621582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,16384,0.27911199463738334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,51200,0.6649733119540745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,51200,0.3261431058247884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,16384,0.2211386627621121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,65536,0.8598471217685275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,16384,0.11019200086593628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,12288,0.212647106912401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,12288,0.16132799784342447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,12288,0.08314310842090182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,10240,0.18335021866692436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,16384,0.248522679011027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,1024,0.02555911077393426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,12288,0.1882151100370619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,8192,0.16665155357784694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,10240,0.06988800234264798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,8192,0.11645600530836318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,10240,0.1600053310394287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,7168,0.13365333610110813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,8192,0.1328222221798367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,7168,0.05222310953670078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,51200,0.7847484482659234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,7168,0.09706666734483506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,65536,1.0240631103515625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,7168,0.11258221997155084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,6144,0.08390933275222778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,6144,0.051141334904564746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,8192,0.05743733379575941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,1024,0.01586844523747762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,5120,0.07242044475343493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,4096,0.08556266625722249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,5120,0.08477333519193862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,6144,0.12762844562530518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,3584,0.07231289148330688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,4096,0.033196445968416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,4096,0.06919199890560575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,5120,0.0986328919728597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,3072,0.06319466564390394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,3584,0.02956177790959676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,6144,0.09868977467219035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,3584,0.06169066826502482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,5120,0.038367112477620445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,3072,0.026731555660565693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,3072,0.04569777846336365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,3072,0.05502577622731527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,4096,0.05843377775616116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,2048,0.04622488882806566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,2560,0.023733332753181458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,2560,0.04828000068664551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,2048,0.020046222541067336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,10240,0.14243022600809732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,2048,0.04138311081462436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,2560,0.05552533268928528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,1536,0.027495111028353374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,1024,0.02713777787155575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,1536,0.01680355601840549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,2560,0.03968888852331374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,1024,0.013792888985739814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,768,0.020776889390415616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,1024,0.028152889675564233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,1536,0.03507377703984579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,768,0.012875555290116204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,512,0.01514488955338796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,768,0.024324445260895625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,1536,0.03470666540993585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,512,0.014574221438831754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,512,0.02086399992307027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,1024,0.021078222327762183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,3584,0.053957333167394005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,256,0.013443555268976422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,256,0.01054222219520145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,256,0.017106667160987854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,768,0.01735911104414198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,128,0.01072088877360026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,128,0.015206222732861837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,32,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,64,0.012510221865442065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,256,0.011476444701353708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,2048,0.03387199838956197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,32,0.013411555853154926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,65536,0.38430044386121964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,65536,0.30761512120564777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,64,0.007281777759393056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,65536,0.14243200090196398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,512,0.011789333489206104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,51200,0.23865421613057455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,51200,0.11287555429670547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,128,0.008832000195980072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,16384,0.0809662209616767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,16384,0.0411102208826277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,128,0.012483555409643384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,12288,0.07697333229912652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,51200,0.30017420980665416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,16384,0.08614844746059841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,10240,0.06475289000405206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,12288,0.06086044179068672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,65536,0.3280871179368761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,10240,0.05203555689917671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,8192,0.05273955398135715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,16384,0.09995910856458877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,10240,0.05896266963746813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,8192,0.04416888952255249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,8192,0.02385422256257799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,8192,0.04889066682921516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,7168,0.021319111188252766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,7168,0.03838933176464505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,12288,0.03216800093650818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,7168,0.04362844427426656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,12288,0.06697333521313138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,10240,0.02795733345879449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,5120,0.035290665096706815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,6144,0.03387022349569533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,6144,0.03915999995337592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,4096,0.02962755494647556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,5120,0.029261334074868098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,7168,0.047472887569003634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,5120,0.03466755482885573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,3584,0.02605155607064565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,4096,0.01516711049609714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,4096,0.024561777710914612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,4096,0.029917332861158583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,3072,0.023397333092159692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,51200,0.2599653402964274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,3584,0.0140666663646698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,3584,0.027456000447273254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,6144,0.01919644408755832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,3072,0.012888888518015543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,5120,0.016931555337376065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,2560,0.017264889346228707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,2560,0.012088889049159156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,2560,0.022198221749729578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,6144,0.040849778387281634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,3584,0.02199466692076789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,2048,0.015032889114485847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,1536,0.013425778183672162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,2048,0.011162666810883416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,2560,0.019721777902709115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,1536,0.009000889129108852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,1536,0.012780444489585029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,3072,0.024676443801985845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,1024,0.01073333372672399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,2048,0.01645866698688931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,768,0.008992888861232335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,2048,0.019702222612169053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,1024,0.010397333237859938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,1536,0.0174319992462794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,768,0.01275199982855055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,1024,0.008045333127180735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,1024,0.015088000231319003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,512,0.008013333711359236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,256,0.005029333134492238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,512,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,512,0.011927111281288994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,768,0.009053332938088311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,128,0.004268444246715969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,256,0.00738844441043006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,256,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,256,0.009719111025333405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,64,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,128,0.007094221810499827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,512,0.007650666766696506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,32,0.004591111093759537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,128,0.009700444837411245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,64,0.007320889168315464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,32,0.00796888851457172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,65536,0.2758595678541395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,65536,0.22342578570048013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,65536,0.12749244107140437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,768,0.007640889121426477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,128,0.007346666521496243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,51200,0.17436710993448892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,16384,0.0722364452150133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,51200,0.10126044352849324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,16384,0.038366221719317965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,12288,0.05899822049670749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,16384,0.08419999811384414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,51200,0.21630222267574736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,65536,0.31046666039360893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,10240,0.049732443359163075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,12288,0.044975999328825206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,3072,0.019681778219011094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,51200,0.24143645498487684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,12288,0.06579733557171293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,10240,0.02607822252644433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,10240,0.03872444563441806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,7168,0.03580888774659898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,8192,0.021987555755509272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,10240,0.056814220216539174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,8192,0.0477439992957645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,16384,0.05894577503204346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,6144,0.0307626658015781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,7168,0.019840889506869845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,12288,0.029740442832310993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,6144,0.018241778016090393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,5120,0.026774222652117412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,6144,0.03834311167399088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,5120,0.016221332881185744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,5120,0.021746666895018682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,8192,0.033237334754731916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,5120,0.033477332856920027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,7168,0.028932445579104956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,4096,0.014698666003015308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,4096,0.018750222192870248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,4096,0.028934223784340754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,3584,0.016861332787407767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,3584,0.01349688900841607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,3072,0.01755911111831665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,3584,0.026760000321600173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,3072,0.012679110798570844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,3072,0.015184000134468079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,2560,0.014727110664049784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,3072,0.023792889383104112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,3584,0.019606222709019978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,2560,0.01384800010257297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,2048,0.012975999878512489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,2560,0.02176355487770504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,7168,0.043614221943749316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,2048,0.00977955593003167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,1536,0.010846222440401712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,2048,0.01926933394538032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,1536,0.008718222379684448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,1024,0.008834666675991481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,1536,0.017264889346228707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,2560,0.011619555453459421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,1024,0.014241778188281588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,768,0.007347555624114142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,6144,0.025734222597546045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,8192,0.041430221663581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,2048,0.012449777788586087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,512,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,768,0.0075528886583116316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,768,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,1536,0.010646222366227044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,768,0.012463111016485425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,256,0.004567999806669024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,512,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,512,0.011728888584507836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,256,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,256,0.00666222224632899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,256,0.01039644413524204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,1024,0.008012444608741337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,128,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,64,0.0038924444880750445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,128,0.009754666851626502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,32,0.003903999924659729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,4096,0.02350933353106181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,64,0.006709333509206772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,512,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,32,0.007035555938879649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,128,0.0038808890514903595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,1024,0.00847111145655314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,65536,0.2086320055855645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,51200,0.21505244572957358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,128,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,65536,0.12447199556562637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,16384,0.07060711251364814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,51200,0.16242044501834446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,51200,0.0990826686223348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,12288,0.057562669118245445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,16384,0.036545776658587985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,16384,0.0843306647406684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,65536,0.3078835540347629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,12288,0.04301688737339444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,12288,0.02869422237078349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,12288,0.06451111369662814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,8192,0.03899911046028137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,10240,0.02495555579662323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,10240,0.05615288681454129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,8192,0.031707555055618286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,16384,0.05596977472305298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,8192,0.04718755682309469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,65536,0.26376178529527455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,10240,0.04604444570011563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,51200,0.24125154813130698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,6144,0.029806223180558946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,10240,0.03710666630003188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,6144,0.023918223049905565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,6144,0.01759733259677887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,6144,0.03740088807211982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,8192,0.021347555849287245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,5120,0.01588444411754608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,4096,0.021846221552954778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,7168,0.027458666099442378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,7168,0.019528888993793063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,4096,0.017868444323539734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,4096,0.014224888549910652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,3584,0.019241778386963736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,4096,0.028498666154013738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,5120,0.02568088968594869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,3072,0.01718488832314809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,3584,0.013146667016877068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,3584,0.01651022169325087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,3584,0.02590399980545044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,2560,0.01480977733929952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,3072,0.012180444267061023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,3072,0.014936000108718872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,3072,0.023741333021057978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,7168,0.04250577754444546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,2560,0.013433777623706393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,2560,0.021073778470357258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,2048,0.011864888999197217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,2048,0.009015999734401703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,1536,0.010598222414652506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,7168,0.034252444903055825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,1536,0.008363555702898238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,1024,0.00868888861603207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,1536,0.016855110724767048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,2560,0.010382222632567087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,1024,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,1024,0.007672888537247975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,2048,0.019084443648656208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,768,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,5120,0.02106933295726776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,5120,0.0329440004295773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,768,0.012173333101802401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,2048,0.012517333030700684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,512,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,768,0.0069475554757648045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,256,0.00664444433318244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,1536,0.010421333213647207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,512,0.005720888988839255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,128,0.0042151109211974675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,256,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,256,0.009500444763236577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,128,0.00675022229552269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,512,0.00701422244310379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,128,0.009451555709044138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,32,0.003911111089918348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,64,0.006650666809744305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,768,0.007324444750944774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,65536,0.19493867291344535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,65536,0.1651217805014716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,128,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,51200,0.15413867102728948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,65536,0.07686222261852689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,32,0.006746666712893381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,65536,0.1729217767715454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,51200,0.12882844607035318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,16384,0.053800887531704374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,51200,0.06127466758092245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,16384,0.045129779312345714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,16384,0.024082667297787134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,16384,0.0479928884241316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,12288,0.04147466686036851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,12288,0.03501866592301263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,12288,0.038667556312349104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,10240,0.033971554703182645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,12288,0.01943911115328471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,256,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,10240,0.03012266755104065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,10240,0.03382222188843621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,10240,0.01721333298418257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,8192,0.025454221500290766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,8192,0.02861600120862325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,7168,0.02541333271397485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,8192,0.015208888385030957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,1024,0.013178666432698568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,7168,0.022375111778577168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,6144,0.022337777747048274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,7168,0.014074666632546319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,6144,0.019627556204795837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,6144,0.02383288906680213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,5120,0.019327999817000497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,6144,0.012950222525331708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,5120,0.01184266640080346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,5120,0.01738755570517646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,51200,0.13291911284128824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,5120,0.021591999464564856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,8192,0.028904888365003798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,3584,0.014526221487257214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,4096,0.010776888993051318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,4096,0.01905777719285753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,3584,0.01391377713945177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,3072,0.013120888835854001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,3584,0.01796533332930671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,512,0.011186666786670685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,3072,0.012930666406949362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,2560,0.01144177797767851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,7168,0.026455110973782007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,4096,0.01516533394654592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,2560,0.011680000358157687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,2560,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,2048,0.010115555591053432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,3584,0.010026666853162978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,2048,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,1536,0.00831200016869439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,2048,0.013780444032616086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,3072,0.016488000750541687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,3072,0.008985777695973715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,1024,0.006601777755551868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,1536,0.008897778060701158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,1536,0.012097777591811286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,2560,0.01517955627706316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,1024,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,768,0.006245333287451003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,1024,0.010108444425794814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,2048,0.01036000003417333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,512,0.00434933313065105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,768,0.006787555499209299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,768,0.009727111293209923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,512,0.006319111006127463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,1536,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,256,0.0038773334688610504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,1024,0.0074746666683091064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,256,0.006174222048785951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,256,0.00832799987660514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,256,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,768,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,64,0.003912888881233004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,128,0.006255110932721033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,128,0.005984888722499211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,32,0.0038897775941424896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,512,0.006326222171386083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,512,0.009388444324334463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,64,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,65536,0.17999022536807588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,32,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,128,0.003552000141806073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,65536,0.15979111194610596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,65536,0.07272889216740926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,4096,0.016404444972674053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,64,0.0038808890514903595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,51200,0.12456266085306804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,51200,0.05856533182991875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,65536,0.15722666846381292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,12288,0.03835466504096985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,16384,0.043967998690075345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,16384,0.023074666659037273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,51200,0.12235732873280843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,16384,0.0477804442246755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,12288,0.018941332896550495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,12288,0.034120000070995755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,12288,0.037280887365341187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,10240,0.01662400033738878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,16384,0.048844443427191846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,10240,0.03310311171743605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,128,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,8192,0.024570667081409033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,8192,0.014583110809326172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,8192,0.028068443139394123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,10240,0.031884445084465876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,7168,0.013558222187889947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,51200,0.13961866166856554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,7168,0.02590311070283254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,8192,0.027648000253571406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,6144,0.01941244469748603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,6144,0.012627555264366997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,5120,0.01809866726398468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,7168,0.024187554915746052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,5120,0.011669333610269757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,4096,0.01551466683546702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,5120,0.020950223008791607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,6144,0.021078222327762183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,4096,0.014890667464998035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,4096,0.010394666757848527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,3584,0.013702222042613559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,6144,0.023606222536828782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,3584,0.013739555246300168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,3072,0.012431999875439538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,3584,0.017856000198258292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,5120,0.016979555288950603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,7168,0.021709332863489788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,3072,0.009026666482289633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,3072,0.012794666820102267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,3072,0.016120889120631747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,2048,0.009393778112199571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,2560,0.008001777860853408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,4096,0.01882755590809716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,2560,0.014944000376595391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,1536,0.008024888734022776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,2048,0.007631111476156447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,2048,0.013091555900043912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,3584,0.009399111072222391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,1024,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,1536,0.0070257774657673305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,1536,0.011792889071835412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,2560,0.011194667054547204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,768,0.005129777722888523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,1024,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,1024,0.007324444750944774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,10240,0.029037333197063867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,1024,0.010957332948843638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,512,0.004260444392760594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,768,0.006353777729802662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,768,0.006710222197903528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,256,0.003899555653333664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,2048,0.010144888526863521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,512,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,512,0.009357333183288574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,1536,0.00869688888390859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,256,0.008763555851247575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,128,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,2560,0.011740444435013665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,128,0.008548444343937768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,64,0.0035751110149754416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,768,0.009711110757456886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,512,0.0063279999627007385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,64,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,65536,0.17419021659427217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,256,0.006248000181383557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,65536,0.12130933337741429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,65536,0.06995555427339342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,51200,0.13614400227864584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,128,0.006246222390068903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,32,0.003527111063400904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,51200,0.09523555305269028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,16384,0.044844445255067616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,51200,0.05669777923160129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,32,0.006365333166387346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,16384,0.02164177762137519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,16384,0.034675555096732244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,16384,0.04422044422891405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,12288,0.026477333572175767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,12288,0.018231110440360177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,10240,0.03138488862249587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,65536,0.1456835534837511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,10240,0.022488888767030504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,8192,0.02482755482196808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,256,0.005988444305128521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,51200,0.1158391104804145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,8192,0.019536000159051683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,8192,0.014242667290899487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,7168,0.021723555194007024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,8192,0.02637333340115017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,12288,0.03497066762712266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,6144,0.01976711054642995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,7168,0.013217777841620974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,7168,0.02415200074513753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,10240,0.01627111103799608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,6144,0.012287110918098025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,128,0.003842666745185852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,12288,0.03464444478352865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,5120,0.011266666981909009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,5120,0.013963555296262106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,4096,0.014455111490355598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,5120,0.019623999794324238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,7168,0.01737155516942342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,4096,0.012529777983824411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,4096,0.017878222796652053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,10240,0.03052088949415419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,6144,0.015555555621782938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,3584,0.011517333487669626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,3584,0.00869333330127928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,3584,0.016513778103722464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,3072,0.011024889018800525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,3072,0.01516533394654592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,4096,0.009704888694816166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,3072,0.008638222184446123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,2560,0.009700444837411245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,6144,0.022103110949198406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,3072,0.011382222175598145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,2560,0.008006222546100616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,2048,0.008376888930797577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,2048,0.013112000293201871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,2560,0.01038755559258991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,2048,0.007370666497283512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,5120,0.016930666234758165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,2560,0.014127110441525778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,2048,0.009019555317031013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,1536,0.00701155596309238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,1024,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,768,0.005241777747869492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,1024,0.006640000061856375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,1536,0.007648888561460707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,3584,0.0125946667459276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,512,0.004570666700601578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,768,0.0064071110553211635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,768,0.006264888991912206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,768,0.00905688852071762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,256,0.003850666599141227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,1024,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,512,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,256,0.00564533347884814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,128,0.003581333491537306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,256,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,1024,0.010075555907355415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,128,0.006005333529578314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,64,0.0032311110860771606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,1536,0.007813333637184566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,32,0.003198222153716617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,64,0.00564444437623024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,256,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,65536,0.1466133329603407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,128,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,128,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,1536,0.011078221930397881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,65536,0.067230224609375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,32,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,51200,0.09348711040284897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,16384,0.04567911227544149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,51200,0.11303644710116917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,65536,0.118541333410475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,16384,0.03336888882848952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,16384,0.02097333305411869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,512,0.0063164445261160536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,65536,0.14272711012098524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,12288,0.01755466726091173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,12288,0.025443555580245122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,10240,0.030831111801995173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,12288,0.03420799970626831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,51200,0.05429066552056206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,10240,0.01555377741654714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,10240,0.021835555632909138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,10240,0.030299554268519085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,12288,0.03621600071589152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,8192,0.014109333356221518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,7168,0.022690667046440974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,8192,0.026132444540659588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,7168,0.013019555144839816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,6144,0.019638222124841478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,7168,0.023744889431529578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,8192,0.025664889150195654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,6144,0.015455111861228943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,5120,0.016795555750528973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,6144,0.011727111207114326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,8192,0.019098665979173448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,512,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,4096,0.014515555567211576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,5120,0.010390222072601318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,5120,0.01924622224436866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,7168,0.017222222354676988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,4096,0.01219377749496036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,3584,0.012795555922720166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,4096,0.017316444052590262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,6144,0.02156177825397915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,3584,0.011391999820868174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,51200,0.12006221877204047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,5120,0.013620444469981723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,3072,0.01039644413524204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,2560,0.010064000056849586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,3072,0.015038222074508667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,4096,0.009392889009581672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,2560,0.007673777639865875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,2560,0.009010666774378883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,2048,0.008776889079146916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,2560,0.013717333475748697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,3584,0.00870488915178511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,3584,0.01647822227742937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,16384,0.043046222792731396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,2048,0.007547555698288812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,2048,0.012096888489193387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,1536,0.007000000112586551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,1024,0.005889777921968036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,3072,0.008044444852405125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,1024,0.006604444649484422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,1024,0.009712888962692684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,768,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,1536,0.007847111258241866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,2048,0.008319999608728621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,768,0.006251555350091722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,512,0.0041999999019834726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,768,0.009027555584907532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,1536,0.01108444482088089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,512,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,256,0.003564444267087512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,512,0.008655110994974772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,1024,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,3072,0.011606222225560082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,256,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,256,0.008023111356629265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,768,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,1536,0.007341333561473423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,128,0.00544177782204416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,128,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,32,0.003536000019974179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,512,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,64,0.00563822231358952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,65536,0.10628533363342285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,32,0.005607999861240387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,51200,0.09477066993713379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,65536,0.11095466878679063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,65536,0.06492800182766385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,65536,0.1416977776421441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,51200,0.0520613325966729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,64,0.003343111111058129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,128,0.00798044436507755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,51200,0.11324089103274876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,12288,0.02978044417169359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,16384,0.03137600090768602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,16384,0.042703999413384326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,12288,0.024439111351966858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,12288,0.016991110311614144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,12288,0.03345511025852627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,51200,0.0876035557852851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,16384,0.035465776920318604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,10240,0.021332444416152105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,8192,0.021930666433440313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,10240,0.02966844373279148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,16384,0.020433778564135235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,8192,0.018734221657117207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,8192,0.025575111309687298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,256,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,10240,0.025268443756633337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,7168,0.01688799924320645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,6144,0.016379555066426594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,7168,0.012309333516491784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,7168,0.023405333360036213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,5120,0.01536266671286689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,6144,0.011063999599880643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,6144,0.01516711049609714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,6144,0.021244444780879553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,4096,0.012768000364303589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,7168,0.019317333896954853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,5120,0.019189332922299702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,128,0.0036142224238978494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,3584,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,4096,0.017196445001496207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,3584,0.010952889091438718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,10240,0.015150222513410779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,3584,0.01588444411754608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,3584,0.008435555630260045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,8192,0.013285333083735572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,3072,0.008019555773999956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,2560,0.00796888851457172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,4096,0.011791999969217511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,3072,0.014534221755133735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,2560,0.007687999970383114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,2048,0.007061333292060428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,3072,0.00886933339966668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,5120,0.009904888768990835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,2048,0.008043555749787224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,1536,0.005971555494599872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,2048,0.007371555599901412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,2048,0.012124444047609964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,4096,0.008887111312813228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,1024,0.004581333448489507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,1536,0.0070675553547011475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,1536,0.010767111347781287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,2560,0.013099555340078143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,1024,0.0063973334100511335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,768,0.004229333251714706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,1024,0.006674666785531574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,1024,0.009749333063761393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,512,0.0035751110149754416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,5120,0.013510222236315409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,768,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,768,0.009027555584907532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,3072,0.010319111247857412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,256,0.003527111063400904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,512,0.006115555349323485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,512,0.008343111309740279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,1536,0.0069751110341813825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,256,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,128,0.003206222214632564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,2560,0.008719111482302347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,256,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,64,0.003516444315512975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,128,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,32,0.0035422220826148987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,128,0.008017777568764156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,768,0.0063484443558586975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,65536,0.1294977797402276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,32,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,65536,0.10805866453382705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,65536,0.06400977902942233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,51200,0.10247200065188938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,65536,0.14203466309441462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,128,0.005376889059940974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,51200,0.05113866594102648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,51200,0.08505511283874512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,51200,0.11290666792127822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,16384,0.03052266769938999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,512,0.005912888795137405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,16384,0.020066666934225295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,16384,0.04228533307711283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,256,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,12288,0.024580443898836773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,12288,0.033815112378862165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,16384,0.04006844427850511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,10240,0.015221332510312399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,8192,0.020392000675201416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,64,0.005595555735958948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,10240,0.029567112525304157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,7168,0.018568888306617737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,8192,0.013117333253224691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,8192,0.018259555101394653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,8192,0.025528889563348558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,6144,0.01628888812330034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,7168,0.016197333733240765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,10240,0.024554666545655992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,7168,0.023621333969963923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,5120,0.013810666898886362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,6144,0.010954666468832227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,6144,0.014856000741322836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,6144,0.020991111795107525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,4096,0.013329777452680798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,5120,0.013245333400037555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,5120,0.018912000788582694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,12288,0.027706666125191584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,4096,0.011628444823953839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,4096,0.008672888908121321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,4096,0.01735377808411916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,3584,0.008341333104504479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,3072,0.00849511143234041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,3584,0.015537778536478678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,10240,0.021074665917290583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,5120,0.009683555199040307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,2560,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,3072,0.008056888977686564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,3072,0.014144889182514615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,3584,0.011053333679835001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,2048,0.006825777805513806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,12288,0.016861332787407767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,2560,0.01314755529165268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,3584,0.01039644413524204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,2048,0.007896888587209914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,2048,0.01145866678820716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,3072,0.009860444400045607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,1536,0.010381333529949188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,1024,0.004277333203289243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,2560,0.007663110891977946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,7168,0.01178311142656538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,1024,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,1536,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,1024,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,768,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,768,0.009015999734401703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,512,0.003587555554178026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,2560,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,512,0.005909333212508096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,1024,0.006666666517655055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,768,0.003891555385457145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,512,0.006234666539563074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,1536,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,256,0.008006222546100616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,1536,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,128,0.0052764444715446895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,2048,0.007373332977294922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,256,0.003568888952334722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,512,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,128,0.00600622221827507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,32,0.002963555562827322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,128,0.008000888758235509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,256,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,64,0.005617777920431561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,65536,0.09314666854010688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,768,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,51200,0.07622044616275363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,65536,0.06992444727155897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,65536,0.14173332850138345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,51200,0.054839111036724515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,64,0.003232888877391815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,51200,0.0499875545501709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,51200,0.11226399739583333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,16384,0.020847110284699336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,16384,0.04218399855825636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,16384,0.019944889677895438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,12288,0.016900445024172466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,16384,0.0306000014146169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,12288,0.016632889707883198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,10240,0.018067555295096505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,128,0.0035377778112888336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,65536,0.06294488906860352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,8192,0.015480000111791821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,10240,0.01496266656451755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,10240,0.02926844358444214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,10240,0.014708444476127625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,7168,0.014003554979960123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,8192,0.012440889245933957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,8192,0.013452444639470843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,8192,0.025439111722840205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,7168,0.011407111254003314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,7168,0.012156444291273752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,12288,0.03314755691422357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,256,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,5120,0.011033777561452655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,6144,0.01072355525361167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,6144,0.01106844428512785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,12288,0.02055022286044227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,6144,0.020638222495714825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,4096,0.009687999884287516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,5120,0.009381333159075843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,5120,0.00996977753109402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,3584,0.009177777502271865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,5120,0.018580444984965853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,4096,0.008678221868144141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,4096,0.01623644431432088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,3072,0.008642666869693333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,3584,0.008086221913496653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,3584,0.015261333849694995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,32,0.005664888769388199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,2560,0.007339555356237624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,3072,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,3072,0.013782222237851886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,2560,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,2048,0.006450666735569636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,2560,0.012735111018021902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,4096,0.009026666482289633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,3584,0.008355555435021719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,2048,0.006609777609507243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,2048,0.011459555890825061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,7168,0.023425777753194172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,3072,0.008028444316652086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,1536,0.006237333433495627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,1024,0.004580444345871608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,1536,0.01039644413524204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,2560,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,768,0.003952888978852166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,1024,0.009351111120647853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,2048,0.0075288886825243635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,6144,0.012155555188655853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,512,0.003559111307064692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,768,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,768,0.008678221868144141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,512,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,512,0.005381333331267039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,512,0.008904000123341879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,1024,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,256,0.005011555635266834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,256,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,128,0.0032435556252797446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,768,0.00565244464410676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,128,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,64,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,128,0.008019555773999956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,32,0.0032319999817344877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,64,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,1024,0.006657777975002925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,65536,0.08012977573606703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,32,0.005312888986534542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,256,0.007985778152942657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,1536,0.005588444570700328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,128,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,1536,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,51200,0.05271466573079427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,16384,0.029697779152128432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,51200,0.048650665415657886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,51200,0.11259822050730388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,16384,0.020084444019529555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,12288,0.02402399977048238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,16384,0.041893333196640015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,65536,0.06704088714387682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,65536,0.14184533225165472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,12288,0.015935111376974318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,10240,0.016688888271649677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,12288,0.016551999582184684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,12288,0.03315111001332601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,8192,0.014418666561444601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,10240,0.014648889501889547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,10240,0.02930755416552226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,8192,0.01312888910373052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,8192,0.012264889147546558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,16384,0.019638222124841478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,7168,0.013026666310098437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,8192,0.02498400045765771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,7168,0.011378666592968835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,6144,0.012081777883900536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,256,0.0035502223504914176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,10240,0.014440889159838358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,6144,0.010671111444632212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,6144,0.010076444182131026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,5120,0.010419555836253695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,6144,0.02051466703414917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,51200,0.06052177482181125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,5120,0.009736888938479954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,4096,0.009566222627957663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,5120,0.01868088874551985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,7168,0.011788444386588203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,3584,0.008674666285514832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,4096,0.008723555339707268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,4096,0.008271999657154083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,4096,0.015858666764365304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,3072,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,3584,0.007991111112965478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,3584,0.014819555812411837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,3072,0.007339555356237624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,2560,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,3072,0.013743999931547375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,7168,0.023039999935362074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,2560,0.012758221891191272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,2048,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,2560,0.007641777396202087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,3584,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,1536,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,2048,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,2048,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,2048,0.011744000017642975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,3072,0.008058666355080074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,65536,0.06167822413974338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,1536,0.010728889041476779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,1024,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,1024,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,768,0.003879111260175705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,1024,0.009544889132181803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,768,0.0063386667105886675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,768,0.005608888963858287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,512,0.003937777959638172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,768,0.009330666727489894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,512,0.005976000179847081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,1024,0.004229333251714706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,512,0.005285333428117964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,512,0.008627555436558193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,256,0.006033777776691649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,128,0.0035582222044467926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,128,0.005026666654480828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,1536,0.006952889263629913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,5120,0.009359111388524374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,64,0.0032266666077905228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,128,0.005985777825117111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,32,0.0032097777972618737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,2560,0.006984000404675801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,256,0.007732444339328342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,64,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,32,0.0052986666560173035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,65536,0.06444178024927776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,65536,0.060247997442881264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,51200,0.04858488837877909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,65536,0.14135911729600695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,256,0.0035386664999855887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,16384,0.022721777359644573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,128,0.008036444584528605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,51200,0.050703111622068614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,256,0.005154666801293691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,51200,0.11213866869608562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,12288,0.019077334139082167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,16384,0.01974311139848497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,16384,0.04188088907135857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,12288,0.015676443775494892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,12288,0.016162667009565566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,12288,0.032799111472235784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,1536,0.006215999937719769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,10240,0.01447111037042406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,8192,0.013834666874673633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,10240,0.029129776689741347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,51200,0.047760887278450855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,7168,0.01276000009642707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,8192,0.01274222218328052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,65536,0.06034577555126614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,7168,0.0107342220014996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,10240,0.017154667112562392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,6144,0.01146311147345437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,7168,0.023061333431137934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,6144,0.010041777458455827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,10240,0.013854222165213691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,6144,0.020575111111005146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,8192,0.011761777930789523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,16384,0.019359111785888672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,5120,0.018044445249769423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,5120,0.009462222456932068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,7168,0.011331555744012197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,4096,0.008663111262851292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,4096,0.015509333875444202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,5120,0.010374222364690568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,5120,0.009016888837019602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,8192,0.024992000725534227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,4096,0.009032888544930352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,3072,0.0076657773719893555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,3584,0.008347555167145198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,3584,0.01479644411140018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,4096,0.008268444074524773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,2560,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,3072,0.007977777885066139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,3072,0.01348355578051673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,2560,0.007620444728268518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,2048,0.006001777946949005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,2560,0.012440889245933957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,2048,0.00655733338660664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,2048,0.011405333048767514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,1536,0.005041777673694822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,2048,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,3072,0.007368000017272101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,6144,0.010684444672531553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,1024,0.004269333349333869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,1536,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,2560,0.006983111302057902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,3584,0.00778311077091429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,768,0.0042239998777707415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,1024,0.005625777774386936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,1024,0.009381333159075843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,512,0.003597333199448056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,768,0.006335999816656113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,768,0.005592888842026393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,768,0.00868711123863856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,256,0.0032266666077905228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,512,0.006052444378534953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,3584,0.00831822223133511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,1536,0.006983111302057902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,128,0.0032017777363459268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,256,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,256,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,256,0.00832622249921163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,64,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,128,0.005969777703285217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,128,0.00498933345079422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,32,0.0028702221396896574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,64,0.00499199993080563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,65536,0.05483377642101712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,512,0.005325333525737126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,512,0.00872444444232517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,65536,0.05876444445716011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,1536,0.01071733319097095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,51200,0.044256889157825045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,65536,0.14121954970889622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,16384,0.01867999964290195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,51200,0.048957334624396436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,51200,0.11196533838907878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,128,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,12288,0.01621955633163452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,16384,0.019366222951147292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,32,0.0053084443012873335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,16384,0.04145688811937968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,10240,0.014752000570297241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,65536,0.06194755766126844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,12288,0.015850666496488783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,12288,0.032848887973361544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,8192,0.014314666390419006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,10240,0.01313688854376475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,10240,0.028804444604449805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,8192,0.011881777809725868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,8192,0.011387555963463254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,7168,0.013028444515334235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,16384,0.018569777409235638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,7168,0.011078221930397881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,7168,0.010767111347781287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,12288,0.014390221900410123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,7168,0.022267555197079975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,6144,0.010036444498433007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,1024,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,6144,0.0098213329911232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,51200,0.04661422305636936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,5120,0.009380444056457942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,4096,0.008831111093362173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,6144,0.011389333340856763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,4096,0.00812711069981257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,10240,0.014186667071448432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,5120,0.01036977767944336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,4096,0.015831111205948722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,5120,0.0090506664580769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,3584,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,5120,0.01755911111831665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,3072,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,3584,0.008362666600280339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,3072,0.008012444608741337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,3072,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,2560,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,3072,0.013170666992664337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,4096,0.008728889127572378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,2048,0.005576888720194499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,2560,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,2560,0.012080888781282636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,8192,0.02404266595840454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,2048,0.007335111498832703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,1536,0.004939555707905027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,2048,0.011448889142937131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,1536,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,1536,0.010741333166758219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,1024,0.004237333519591225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,3584,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,6144,0.019768888751665752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,1024,0.005631111148330901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,768,0.0038782221575578055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,1024,0.009338666995366415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,3584,0.014797333214018079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,768,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,768,0.005611555443869696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,768,0.009315555294354757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,1536,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,2560,0.007650666766696506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,256,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,512,0.006237333433495627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,512,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,1024,0.00665866666369968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,2048,0.006673777682913675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,256,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,256,0.00795555528667238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,512,0.00360000009338061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,128,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,64,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,128,0.007921777665615082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,65536,0.04003289010789659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,512,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,64,0.004951999833186467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,32,0.005112888912359874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,51200,0.030956443813112047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,65536,0.05730933613247342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,128,0.002879111096262932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,65536,0.14112267229292128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,51200,0.04672355453173319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,128,0.00500177757607566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,51200,0.11163466506534153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,32,0.002847111059559716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,12288,0.013257777525318993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,16384,0.016906667086813185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,16384,0.041433778074052595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,12288,0.015219555960761176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,65536,0.059169775909847684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,10240,0.012636444634861417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,12288,0.03223022156291538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,10240,0.012879999975363413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,10240,0.013105777402718862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,10240,0.028625776370366413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,16384,0.018616000811258953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,8192,0.011395555403497485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,7168,0.01072799993885888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,8192,0.011395555403497485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,12288,0.014212444424629211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,7168,0.010475555227862464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,6144,0.010043555663691627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,7168,0.021635555558734473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,7168,0.010778666370444827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,8192,0.011397333608733283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,5120,0.009482666850090027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,6144,0.010032888915803697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,6144,0.009707555174827576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,6144,0.01925066610177358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,4096,0.00796977761718962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,5120,0.009368889033794403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,8192,0.024033778243594702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,5120,0.017820444371965196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,3584,0.007344889144102733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,4096,0.008664000365469191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,4096,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,4096,0.015508444772826301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,3072,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,51200,0.04618311259481642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,3584,0.008185777399275038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,3584,0.014820444915029736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,2560,0.005976888868543837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,3072,0.008030222521887885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,3072,0.007325333025720384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,3072,0.01316088851955202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,2048,0.005592888842026393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,2560,0.0069511110583941145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,2560,0.012256888879670037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,256,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,1536,0.004902222090297275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,2048,0.007359111474619971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,2048,0.011439111497667102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,3584,0.00831288927131229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,16384,0.015190222197108798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,1024,0.004230222354332606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,1536,0.006949333681000604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,1536,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,768,0.0038808890514903595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,1024,0.006605333338181178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,1024,0.005633777628342311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,1024,0.009352888498041365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,5120,0.008791111409664154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,512,0.003551111039188173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,768,0.005595555735958948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,512,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,512,0.00850666645500395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,512,0.006054222169849608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,1536,0.01071911139620675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,256,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,2560,0.007598222129874759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,256,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,128,0.0028693332440323303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,256,0.008029333419269985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,768,0.009062222308582729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,128,0.004936000125275718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,128,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,32,0.002842666581273079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,64,0.005220444252093633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,768,65536,0.030392001072565716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,32,0.004895110925038655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,768,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,65536,0.0579013360871209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,65536,0.05702755848566691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,768,51200,0.02495199938615163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,64,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,256,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,768,16384,0.01315111087428199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,51200,0.04598311252064175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,51200,0.11131289270189072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,16384,0.018167111608717177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,16384,0.01686044368478987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,16384,0.04133955637613932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,12288,0.014171555638313293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,12288,0.014499555031458536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,2048,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,12288,0.03189599845144484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,10240,0.012808000048001608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,51200,0.045873777733908765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,768,8192,0.011798222031858234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,10240,0.02791911032464769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,8192,0.01146755533085929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,8192,0.01124533348613315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,8192,0.0239964442120658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,768,10240,0.012808889150619507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,65536,0.14085866345299616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,7168,0.02156977852185567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,7168,0.010703999963071613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,10240,0.012903110848532783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,6144,0.009583111438486312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,6144,0.010056888891590966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,768,5120,0.009127111070685917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,768,12288,0.01146933353609509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,768,7168,0.010770666930410596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,768,4096,0.008404444489214156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,5120,0.009408888717492422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,5120,0.008731555607583787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,5120,0.01720711092154185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,7168,0.010408889088365767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,4096,0.00867377801073922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,4096,0.008043555749787224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,768,6144,0.010029333333174387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,4096,0.015505777464972602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,3584,0.008352888955010308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,3584,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,3584,0.014489778214030795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,768,2560,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,6144,0.019295111298561096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,3072,0.013471110827393003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,2560,0.006990222467316522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,2560,0.007725333174069722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,768,2048,0.005223999834722943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,2560,0.012149333126015134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,768,3584,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,128,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,2048,0.007300444775157505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,768,3072,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,2048,0.011400889191362592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,768,1024,0.003920000046491623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,1536,0.005978666659858491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,3072,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,1024,0.006342222293217977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,768,768,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,1024,0.009594666461149851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,768,1536,0.004912000149488449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,768,0.0052977779673205475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,768,512,0.003552000141806073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,768,0.009016888837019602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,1536,0.00673333348499404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,3072,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,512,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,1024,0.005667555663320754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,256,0.004944889081848992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,256,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,2048,0.0063386667105886675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,768,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,1536,0.010370666782061258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,128,0.008014221986134848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,128,0.005977777971161737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,512,0.008691555923885768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,768,32,0.0029208889851967492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,64,0.004959110998445087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,256,0.005791999813583162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,768,128,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,65536,0.0567493306265937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,65536,0.14025510682000056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,768,256,0.003208000005947219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,512,51200,0.020240000552601285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,512,65536,0.025046222739749487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,51200,0.04544622368282742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,51200,0.044657776753107704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,512,16384,0.013236444857385425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,128,0.005144000053405762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,51200,0.11180444558461507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,512,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,16384,0.016844444804721408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,16384,0.01732711162832048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,16384,0.04043644335534837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,12288,0.014362666341993542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,12288,0.0314008891582489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,10240,0.012758221891191272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,768,64,0.00282222218811512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,512,12288,0.012039999994966718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,512,8192,0.009346666435400644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,10240,0.012936000194814471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,10240,0.02815199891726176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,512,7168,0.00888533310757743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,12288,0.014455111490355598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,8192,0.011420444481902652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,512,10240,0.010770666930410596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,512,6144,0.008115555677149031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,7168,0.010727110836240979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,7168,0.010378667049937779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,7168,0.02163644466135237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,512,5120,0.007683555285135905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,6144,0.009755555954244401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,65536,0.05705600314670139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,6144,0.019103111492262948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,512,4096,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,5120,0.009379555781682333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,5120,0.008718222379684448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,5120,0.017285333739386666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,512,3584,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,4096,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,4096,0.015264888604482016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,3584,0.007679999702506595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,512,3072,0.006657777975002925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,6144,0.010048000348938836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,3072,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,3072,0.013168888787428537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,512,2560,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,3072,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,8192,0.01126844435930252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,8192,0.023369777533743117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,512,2048,0.005416000054942236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,2560,0.007607999775144789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,3584,0.008310221963458592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,3584,0.014821334017647637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,512,1536,0.004566222015354368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,2048,0.00702400008837382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,2048,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,2048,0.011304000185595619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,512,1024,0.0038728887836138406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,1536,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,1536,0.005963555640644497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,1536,0.010065777434243096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,32,0.005008888741334279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,512,768,0.0035422220826148987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,1024,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,1024,0.009703111317422655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,512,512,0.0032026666320032547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,768,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,768,0.005630222045713001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,768,0.008752000000741746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,512,256,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,512,0.006256888724035687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,512,0.0052800000541739995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,512,0.008034666379292807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,512,128,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,256,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,256,0.004941333499219683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,256,0.008026666939258575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,4096,0.008671110702885522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,1024,0.006612444503439798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,128,0.004967111266321606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,128,0.008023999631404877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,256,65536,0.020620443754725985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,64,0.00490399988161193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,32,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,65536,0.05548977851867676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,256,51200,0.017616889543003507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,65536,0.140775998433431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,128,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,51200,0.044510222143597074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,512,32,0.002846222163902389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,256,16384,0.010039110978444418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,51200,0.040136890278922185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,2560,0.006715555571847492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,256,12288,0.008667555948098501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,16384,0.016598222984208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,16384,0.04009244508213467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,65536,0.04974666568968031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,12288,0.014127999544143677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,256,10240,0.009369778136412302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,12288,0.014127110441525778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,12288,0.03143644332885742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,10240,0.013058666553762225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,512,64,0.0028435554769304064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,10240,0.02752444479200575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,51200,0.11150756147172715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,2560,0.012454222473833295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,8192,0.0110880004035102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,256,7168,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,8192,0.023357333408461675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,7168,0.01073155552148819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,7168,0.010402667025725046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,256,6144,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,7168,0.021695110532972548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,10240,0.012683555483818054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,6144,0.010092444717884064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,256,5120,0.008220444122950235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,6144,0.01896711190541585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,5120,0.00869333330127928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,8192,0.011414222419261932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,256,4096,0.007680888805124495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,5120,0.01721333298418257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,4096,0.00867733359336853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,4096,0.007979555262459649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,4096,0.01577333278126187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,256,8192,0.008355555435021719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,3584,0.007603555917739868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,256,3072,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,3584,0.00833155545923445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,3584,0.014527999692493014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,256,2560,0.005968889014588461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,16384,0.017156443662113614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,3072,0.006950221955776215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,3072,0.013447110851605734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,256,2048,0.005081777771313985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,2560,0.007648000286685095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,2560,0.012440000143316058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,2048,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,256,1536,0.004551110996140374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,6144,0.009408888717492422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,2048,0.01108888867828581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,5120,0.009316444396972656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,1536,0.005923555543025334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,256,1024,0.0038746665749284956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,1536,0.006647999915811751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,2560,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,1024,0.005608888963858287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,1024,0.00904622260067198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,2048,0.007247111035717859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,768,0.008667555948098501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,256,512,0.0031866667171319327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,1536,0.01069866700304879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,256,3584,0.007017778025733099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,256,768,0.0035946667194366455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,256,256,0.0031902222997612427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,512,0.005994666781690385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,768,0.005587555468082428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,256,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,256,128,0.0028906667398081887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,256,0.008114666574531132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,3072,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,128,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,128,0.004972444640265571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,128,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,256,32,0.0026888888743188647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,512,0.008584000170230865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,64,0.004976888911591636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,32,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,768,0.006315555423498154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,65536,0.054358220762676664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,128,51200,0.01370488852262497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,65536,0.049308445718553334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,65536,0.13967377609676784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,256,64,0.002569777684079276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,512,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,1024,0.006508444332414203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,51200,0.042846222718556724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,51200,0.1099555558628506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,128,65536,0.015203555425008139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,16384,0.016912000046836007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,16384,0.016910221841600206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,16384,0.03975911272896661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,128,10240,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,12288,0.014225777652528552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,12288,0.031425777408811785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,10240,0.012436444560686747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,10240,0.012977778083748288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,128,8192,0.00832088871134652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,51200,0.03973955578274197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,8192,0.011407111254003314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,8192,0.011130666567219628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,128,7168,0.008035555481910706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,8192,0.023424888650576275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,256,0.005919999960396025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,7168,0.010573333336247338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,7168,0.010388444695207808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,7168,0.02130488885773553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,6144,0.009131555755933126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,128,16384,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,128,5120,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,10240,0.02750933335887061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,5120,0.009404444032245213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,5120,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,128,4096,0.006724444528420766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,12288,0.013976888524161445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,128,6144,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,4096,0.007836444510353936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,4096,0.008626666333940294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,128,3584,0.006816888848940532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,6144,0.010043555663691627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,6144,0.019312888383865356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,128,3072,0.006312888943486744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,3584,0.008175111479229396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,3584,0.014468444718254937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,128,12288,0.008340444829728868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,128,2560,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,5120,0.01720355616675483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,3072,0.01308444473478529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,2560,0.006800888727108638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,128,2048,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,2560,0.012107555237081317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,3584,0.007600888609886169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,2048,0.006331555545330048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,128,1536,0.004272888931963178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,3072,0.008010666403505538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,1536,0.005973333285914526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,1536,0.010553778045707278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,2560,0.007654222349325816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,128,1024,0.0038844446341196695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,4096,0.015182221929232279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,1024,0.006575111299753189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,1024,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,1024,0.009709333380063375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,2048,0.01106844428512785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,128,512,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,768,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,768,0.005608888963858287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,768,0.00904622260067198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,3072,0.007342221836249034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,512,0.008384000096056197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,1536,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,256,0.0052328887912962176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,128,128,0.0029066666546795103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,256,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,256,0.008095111283991072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,2048,0.007338666253619724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,128,64,0.002847111059559716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,128,0.004976888911591636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,128,32,0.002716444432735443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,128,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,64,65536,0.014713777436150445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,128,256,0.0028737777223189673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,64,51200,0.01257955531279246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,32,0.005151111218664381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,64,16384,0.009011555876996782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,65536,0.05277599891026815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,64,12288,0.007979555262459649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,51200,0.042542222473356456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,64,10240,0.007367110914654202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,128,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,64,8192,0.006955555743641323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,10240,0.01276266657643848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,512,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,64,7168,0.00666222224632899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,64,6144,0.006315555423498154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,8192,0.011056889262464313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,64,5120,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,7168,0.010038221875826517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,64,4096,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,6144,0.009411555197503831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,16384,0.016931555337376065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,5120,0.008648888932334052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,64,3072,0.006122666514582104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,12288,0.014148443937301636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,64,2560,0.006301333506902059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,3584,0.007603555917739868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,512,0.006029333505365584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,64,2048,0.005976000179847081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,3072,0.006953777538405524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,64,1536,0.004262222184075249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,64,1024,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,2560,0.006726222319735422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,2048,0.0063324446479479475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,1536,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,128,768,0.0035617777870761026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,64,512,0.0032071111102898917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,1024,0.005648889061477449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,64,256,0.002874666617976295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,64,128,0.002883555574549569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,768,0.005566222386227713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,64,64,0.002583111118939188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,512,0.005118222286303838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,256,0.005239999956554837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,128,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,32,65536,0.014479110638300577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,32,0.004979555391603046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,32,51200,0.012505778008037143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,64,768,0.0035502223504914176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,65536,0.05267288949754503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,64,3584,0.006309333360857434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,32,12288,0.007690666450394525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,4096,0.007663110891977946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,16384,0.01697688963678148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,64,32,0.002509333400262727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,64,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,12288,0.014197332991494073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,32,8192,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,32,7168,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,10240,0.012430222498046027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,32,6144,0.006301333506902059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,8192,0.011100444528791638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,32,16384,0.008674666285514832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,7168,0.010414222048388587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,32,4096,0.005984888722499211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,6144,0.009366222553782994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,32,3584,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,5120,0.008637333081828224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,32,3072,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,4096,0.00775911079512702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,32,2560,0.00665511108107037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,3584,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,32,2048,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,3072,0.0069822221994400024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,32,1536,0.004355555607212914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,2560,0.006671111202902264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,32,1024,0.003882666842805015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,2048,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,32,768,0.003539555602603488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,51200,0.04253955682118734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,1024,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,768,0.0052879999081293745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,512,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,32,128,0.0029066666546795103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,256,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,32,64,0.0028400001012616684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,128,0.004983110974232356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,32,5120,0.006302222195598815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,64,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,1536,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,32,0.004961777892377642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,32,256,0.002955555501911375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,64,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,16384,0.38631733258565265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,32,32,0.002567111204067866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,51200,0.8429208861456977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,51200,1.2043742073906791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,32,512,0.0032142222755485107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,12288,0.2924586666954888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,192,32,10240,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,10240,0.23559289508395723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,12288,0.20269332991706002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,51200,0.41524532106187606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,8192,0.1943333281411065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,10240,0.1679662201139662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,12288,0.10508622063530816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,7168,0.17263466782040068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,8192,0.07179644372728136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,16384,0.2721253236134847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,7168,0.0637342201338874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,7168,0.11849333180321588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,5120,0.12301245000627305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,16384,0.14013155301411948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,6144,0.10929866631825765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,4096,0.10445421934127808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,5120,0.04718933171696133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,10240,0.08720711204740737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,3584,0.08976178036795722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,4096,0.07424533367156982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,3072,0.0765377746687995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,8192,0.138410660955641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,3584,0.06456977791256376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,2560,0.06446133057276408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,3072,0.03164711263444688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,3072,0.05496533380614387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,2048,0.051947553952534996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,2560,0.027759111589855615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,2560,0.04814666509628296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,1536,0.040601776705847845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,2048,0.024041778511471216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,4096,0.04006310966279771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,2048,0.04070133301946852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,1024,0.029520889123280842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,1536,0.019375110665957134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,3584,0.03546311126814948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,6144,0.1444257762697008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,1024,0.01571644511487749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,512,0.016732444365819294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,1024,0.025063999825053748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,768,0.014122666584120857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,6144,0.05480800072352091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,256,0.01074488874938753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,512,0.013106666505336761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,128,0.00959822204377916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,5120,0.08650488985909356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,256,0.012075555821259817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,64,0.009649777577983009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,128,0.011730666789743634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,1536,0.03216266632080078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,32,0.010411555568377176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,32,0.01534400052494473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,512,0.018391999933454726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,256,0.015712888704405892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,65536,1.1512426800198026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,768,0.022786666949590046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,65536,0.8520586755540637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,768,0.024145777026812237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,16384,0.29190222422281903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,51200,0.9071013132731119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,65536,0.420775122112698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,128,0.014872888724009195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,12288,0.23149776458740234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,16384,0.21570755375756157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,51200,0.32477688789367676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,12288,0.1689457760916816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,10240,0.20253333780500624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,10240,0.1415217717488607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,16384,0.10856711202197605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,8192,0.16456088754865858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,10240,0.07010755274030897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,7168,0.145778669251336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,64,0.014498665928840637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,8192,0.05673778057098389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,8192,0.11492355664571126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,6144,0.12860711415608725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,7168,0.0971226692199707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,6144,0.04430222180154589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,6144,0.08521955543094212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,4096,0.08867110808690389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,12288,0.08362755510542129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,5120,0.07098755571577284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,3584,0.07919555240207247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,4096,0.05760177638795641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,3584,0.05111644334263272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,51200,0.6605084207322862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,5120,0.10698844326866998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,3072,0.026406221919589575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,3072,0.0451546675629086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,5120,0.03817866577042474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,2560,0.023447111248970032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,2048,0.04734311170048184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,4096,0.03279555506176419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,1536,0.0373946660094791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,2048,0.019831111033757527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,2048,0.033300442828072443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,3584,0.029260443316565618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,1024,0.02812622321976556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,1536,0.016493333710564505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,768,0.02166844407717387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,1024,0.020745777421527438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,768,0.012803555362754397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,2560,0.03935022155443827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,3072,0.06759822368621826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,512,0.015211555692884656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,256,0.01167555567291048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,1024,0.013814222481515674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,7168,0.051009777519438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,256,0.010478222535716163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,256,0.013482666677898832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,768,0.020305777589480083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,512,0.01616533266173469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,64,0.006270222365856171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,128,0.01072799993885888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,128,0.012780444489585029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,64,0.012467555701732635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,32,0.013129777378506131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,65536,0.37550844086541074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,512,0.011797332929240333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,2560,0.05733333031336466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,128,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,65536,0.30453332265218097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,16384,0.09574311309390598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,65536,0.1397635539372762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,51200,0.23481867048475477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,32,0.006648889018429651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,12288,0.07559555768966675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,51200,0.11083910862604777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,16384,0.040255109469095864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,10240,0.062401778168148465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,12288,0.03199555476506551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,8192,0.051389333274629384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,10240,0.05134310987260607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,8192,0.023767999476856653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,8192,0.043430222405327686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,51200,0.2934817738003201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,7168,0.021360889077186584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,6144,0.03985333442687988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,16384,0.07905689213011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,6144,0.018925334016482036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,10240,0.02755911151568095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,5120,0.0340595543384552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,6144,0.033407111962636314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,4096,0.02981688910060459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,7168,0.0455644428730011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,5120,0.017048888736300997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,3584,0.025423111187087163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,4096,0.01512977812025282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,4096,0.02432800001568264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,3072,0.02234222160445319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,7168,0.037978665696250066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,3584,0.02209866709179348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,2560,0.01923644377125634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,3072,0.013003555436929068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,3072,0.01943288909064399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,2560,0.012101333174440594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,2048,0.016152888536453247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,2560,0.01688799924320645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,1536,0.012758221891191272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,2048,0.011104889214038849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,2048,0.01481244464715322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,1024,0.009947555760542551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,5120,0.028797331783506606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,1536,0.012700444294346703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,768,0.008784000244405534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,3584,0.01387111097574234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,1024,0.010663111176755695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,512,0.006680888848172293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,768,0.007707555260923173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,768,0.009743111001120673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,12288,0.05991466840108236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,512,0.008352888955010308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,1536,0.02699111070897844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,256,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,128,0.0042168887125121225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,128,0.006983111302057902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,64,0.004271999829345279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,256,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,32,0.00423200014564726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,256,0.007516444557242923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,1536,0.009032888544930352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,1024,0.008347555167145198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,128,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,65536,0.2170382208294339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,512,0.007409777906205919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,51200,0.16889422469668916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,16384,0.07212089167700873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,51200,0.10087200005849202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,12288,0.05868177943759494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,32,0.007887110941939885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,16384,0.03808355662557814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,16384,0.05805777841144138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,10240,0.04682577649752299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,65536,0.29004356596204967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,12288,0.0299582216474745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,8192,0.04107466671201918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,10240,0.02572533322705163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,51200,0.21771377987331816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,10240,0.038146668010287814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,8192,0.021910222040282354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,8192,0.032812443044450544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,6144,0.030741333961486816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,7168,0.01994222237004174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,5120,0.026587555805842083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,6144,0.018050667312410142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,6144,0.025426665941874187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,4096,0.023378666904237535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,5120,0.021648888786633808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,12288,0.044347554445266724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,3584,0.01982933282852173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,7168,0.03521777855025397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,64,0.007328888608349695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,3584,0.013520888984203339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,7168,0.028561777538723413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,3072,0.012721777790122561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,2560,0.014800000521871777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,4096,0.014754666222466363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,2560,0.013694222602579327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,2048,0.012996444271670448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,2560,0.01167022188504537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,3072,0.01758044461409251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,1536,0.010947555303573608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,3584,0.01658933361371358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,2048,0.012323555847009024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,1024,0.009000000026490953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,1536,0.0086986662613021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,1536,0.010606221854686737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,768,0.007620444728268518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,1024,0.00867022242810991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,4096,0.01851377718978458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,512,0.006495999793211619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,768,0.008303110798199972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,512,0.007265778051482306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,2048,0.009763555394278632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,256,0.004902222090297275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,512,0.007248000138335758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,128,0.004587555511130227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,256,0.006983111302057902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,256,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,64,0.003887999802827835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,128,0.006663110935025745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,32,0.004247111164861255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,768,0.007613333563009898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,64,0.006653333289755716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,32,0.0070062221752272705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,65536,0.26507022645738393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,5120,0.016164445214801364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,65536,0.20317422019110787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,65536,0.12302222516801621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,128,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,65536,0.12736533747778997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,51200,0.09879822201199001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,12288,0.05548266569773356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,16384,0.055162668228149414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,16384,0.036433776219685875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,12288,0.028943111499150593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,10240,0.046682665745417275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,51200,0.20583111710018584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,1024,0.008060444560315874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,8192,0.03865689039230347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,10240,0.024591111474566992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,51200,0.1596791081958347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,7168,0.03449155555831061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,8192,0.021288888321982488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,8192,0.03126310971048143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,6144,0.029513776302337646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,7168,0.019303111566437613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,7168,0.02698933415942722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,5120,0.02576444380813175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,3072,0.015192889504962497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,6144,0.023981332778930664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,4096,0.021662222014533147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,5120,0.020590222544140287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,10240,0.03661600086424086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,3584,0.01927911076280806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,4096,0.014352889524565803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,4096,0.01761866609255473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,3072,0.016723554995324876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,3584,0.013100444442696042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,3584,0.016255110502243042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,12288,0.04265422291225857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,3072,0.014848889576064216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,2560,0.010868444210953183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,2048,0.012857777376969656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,2048,0.009012444151772393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,1536,0.010834666589895884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,2048,0.011791110866599612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,1536,0.008446222378147973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,1024,0.00871377769443724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,3072,0.012096888489193387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,6144,0.01754488878779941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,768,0.0069546666410234236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,1024,0.008256889051861232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,16384,0.07394666804207696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,512,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,768,0.007325333025720384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,768,0.008006222546100616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,256,0.004783111313978831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,512,0.0070186663005087115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,512,0.0069466663731469055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,128,0.004046222401989831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,256,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,256,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,2560,0.014766222900814481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,64,0.004219555606444676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,32,0.004558222161398994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,128,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,5120,0.015846222639083862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,64,0.006657777975002925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,65536,0.21994488769107393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,32,0.00664444433318244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,51200,0.17170133855607775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,65536,0.07610399855507745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,65536,0.16198578145768908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,16384,0.06101422177420723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,51200,0.12662400139702693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,1536,0.010207110808955299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,12288,0.04636711213323805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,128,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,16384,0.04422666629155477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,12288,0.019352000620630052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,12288,0.03445333242416382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,8192,0.0322053333123525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,10240,0.01703733371363746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,10240,0.02948088778389825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,7168,0.028295109669367473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,8192,0.024714666936132643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,7168,0.013968000809351603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,7168,0.021692444880803425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,6144,0.02474755545457204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,16384,0.02385155525472429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,5120,0.021670222282409668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,6144,0.012824888858530255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,2560,0.013242666920026144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,6144,0.01940000057220459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,4096,0.017901332841979135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,5120,0.011911999848153857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,3584,0.01591022147072686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,4096,0.010835555692513784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,8192,0.01533066729704539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,4096,0.014933332800865173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,3584,0.009697777529557547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,3584,0.013824889229403602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,2560,0.012437333663304647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,3072,0.009039999710188972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,3072,0.012812444733248817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,10240,0.03818133473396301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,2048,0.011043555206722684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,2560,0.011578666667143503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,1536,0.009348444640636444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,2048,0.010371555884679159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,3072,0.014478221535682678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,1536,0.007343110938866933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,1024,0.007287999822033777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,1536,0.00868000007337994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,51200,0.061030222309960254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,768,0.00572622236278322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,1024,0.007868444754017724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,768,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,512,0.004576888763242298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,768,0.007272889216740926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,2048,0.007674666742483775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,256,0.0042408886882993905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,512,0.006325333482689328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,512,0.006331555545330048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,128,0.003531555748648114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,256,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,1024,0.006903111106819577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,128,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,128,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,32,0.003519111209445529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,64,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,32,0.006613333192136552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,65536,0.20536443922254774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,256,0.006254222244024277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,65536,0.15663733747270372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,51200,0.1607475545671251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,65536,0.07185066408581205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,2560,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,16384,0.055028445190853544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,51200,0.05829333596759372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,12288,0.043395555681652494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,16384,0.04321066538492838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,10240,0.0362044440375434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,12288,0.018619555566045973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,12288,0.033365332418017916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,8192,0.031199110878838435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,10240,0.02846933404604594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,1024,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,64,0.0035413333939181436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,8192,0.014840000205569796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,7168,0.013515555196338229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,16384,0.02276888820860121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,6144,0.02379200028048621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,6144,0.012653333445390066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,5120,0.02036977807680766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,6144,0.019129777948061626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,10240,0.01651822196112739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,4096,0.016928889685206942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,5120,0.011618666350841522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,5120,0.016864889197879367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,8192,0.024153777294688757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,4096,0.01482222146458096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,51200,0.12222577465905084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,5120,0.017280000779363845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,7168,0.021300445000330608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,3584,0.013458666702111563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,2560,0.012074666718641916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,3072,0.008691555923885768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,3072,0.012773333324326409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,2560,0.008004444340864817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,2560,0.011415111521879831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,7168,0.02700533303949568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,2048,0.007355555891990662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,3584,0.015145777000321282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,1536,0.007295110987292395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,3584,0.009293333523803288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,1024,0.007289778027269576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,2048,0.01072266697883606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,768,0.005471999860472149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,1024,0.006959999601046245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,1536,0.009011555876996782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,512,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,3072,0.013783111340469785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,512,0.006351111249791251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,512,0.006355555521117316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,256,0.005912888795137405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,4096,0.01014666673209932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,128,0.003528000166018804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,1024,0.006789333290523953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,128,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,128,0.005998222364319696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,64,0.00388355553150177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,32,0.0038924444880750445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,64,0.006315555423498154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,32,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,2048,0.010041777458455827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,65536,0.17920888794793022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,768,0.006991110742092133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,51200,0.14135911729600695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,1536,0.008350222474998897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,65536,0.06951822174919976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,65536,0.11904799938201904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,256,0.004212444441186057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,12288,0.03579644362131754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,16384,0.03387644555833604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,12288,0.01822933389080895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,12288,0.02597155504756504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,10240,0.031071109904183283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,768,0.0064702220261096954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,8192,0.02653955585426754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,10240,0.016187555260128446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,10240,0.022240888741281297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,7168,0.02253244486120012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,8192,0.014422222971916199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,51200,0.09295822514428033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,8192,0.019313777486483257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,16384,0.021751999855041504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,6144,0.020611556039916146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,7168,0.013355555633703867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,5120,0.01756177842617035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,6144,0.0124177775449223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,6144,0.015582222077581616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,4096,0.015215999550289579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,256,0.006166222194830577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,5120,0.013903111219406128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,3584,0.013390222357379066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,4096,0.009438222481144799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,4096,0.012483555409643384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,3072,0.012348444097571902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,3584,0.011487111449241638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,3072,0.008381333616044786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,2560,0.010727110836240979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,5120,0.011049778097205691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,2560,0.009677333136399588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,2048,0.010064000056849586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,16384,0.0492142207092709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,3584,0.008664000365469191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,1536,0.008663111262851292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,2048,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,2048,0.008536888493431939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,1024,0.00664533343580034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,1536,0.0070088886552386815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,1536,0.007286222444640265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,768,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,1024,0.006657777975002925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,768,0.006324444380071428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,2560,0.007696888513035244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,768,0.006502222269773483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,256,0.003951999876234266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,512,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,512,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,128,0.003519999898142285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,256,0.005979555762476391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,256,0.005684444473849402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,64,0.0032951111594835916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,128,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,128,0.005585777676767773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,1024,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,64,0.005704888867007361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,512,0.004597333156400257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,32,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,51200,0.1215475532743666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,65536,0.06747821966807048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,65536,0.11718044016096328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,51200,0.056927111413743764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,16384,0.04288088944223192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,51200,0.05429777834150526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,7168,0.01702044407526652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,51200,0.09045244587792291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,12288,0.033942222595214844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,16384,0.020991999242040846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,16384,0.03269511130121019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,12288,0.01754577789041731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,8192,0.02453511125511593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,65536,0.1463822258843316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,10240,0.021801778011851843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,7168,0.02134311033619775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,8192,0.014101333088344999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,8192,0.01889244384235806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,6144,0.018752889500723947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,7168,0.012844444149070315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,7168,0.016565332810084026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,5120,0.015885333220163982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,6144,0.011735111474990845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,6144,0.015142222245534262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,10240,0.028902222712834675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,5120,0.010400888820489248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,5120,0.013696889082590738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,3584,0.01204355557759603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,4096,0.012118221984969245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,3584,0.008546666966544257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,3584,0.01126933346192042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,3072,0.011004444625642566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,32,0.0032435556252797446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,3072,0.010455999937322406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,2560,0.010216889282067617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,3072,0.008030222521887885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,12288,0.024998222788174946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,2560,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,2560,0.00906666699382994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,1536,0.007295110987292395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,2048,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,2048,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,1024,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,1536,0.007330666813585494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,1024,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,4096,0.013531555732091268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,768,0.00498933345079422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,10240,0.015711999601787992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,768,0.0063457778758472866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,4096,0.009085333181752099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,768,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,256,0.003576888806290097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,512,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,512,0.005991999887757831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,3072,0.011418666276666852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,128,0.0034640000926123727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,256,0.006057777752478917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,256,0.005595555735958948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,64,0.0035582222044467926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,1024,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,32,0.003548444559176763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,2048,0.008616000413894653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,64,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,512,0.004280888785918553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,65536,0.11144088374243842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,51200,0.08788800239562988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,65536,0.06492177645365398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,1536,0.0069822221994400024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,16384,0.036432888772752546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,51200,0.05215200119548374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,51200,0.08595199717415704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,12288,0.028967999749713477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,16384,0.03053511182467143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,12288,0.016866667403115165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,32,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,10240,0.023967110448413428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,10240,0.015174221661355762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,8192,0.0199306673473782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,128,0.005990222096443176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,16384,0.02055022286044227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,8192,0.013350221845838757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,7168,0.0173680004146364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,6144,0.015169777803950839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,7168,0.012089778151777057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,12288,0.02386577758524153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,6144,0.011051555474599203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,5120,0.013726222018400827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,10240,0.02129688858985901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,5120,0.013382222089502545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,128,0.00564533347884814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,8192,0.01827555563714769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,4096,0.011750222080283694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,3584,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,7168,0.0162435554795795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,3072,0.009019555317031013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,3584,0.010703110860453712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,65536,0.10746222072177464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,2560,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,3072,0.007984889050324758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,5120,0.009700444837411245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,2048,0.006720888945791457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,2560,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,2560,0.008678221868144141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,1536,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,2048,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,4096,0.008721777962313758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,1024,0.004931555440028508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,1536,0.007121777368916406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,1536,0.007265778051482306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,768,0.00423911131090588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,1024,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,1024,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,3072,0.009703999592198266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,768,0.006314666734801398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,2048,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,512,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,256,0.003536888708670934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,256,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,3584,0.00832177781396442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,256,0.00563466673096021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,6144,0.01484444406297472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,64,0.0031893334041039148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,768,0.006370666540331311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,4096,0.012473777764373355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,64,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,512,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,65536,0.10279911094241673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,128,0.003551111039188173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,65536,0.06374044550789727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,65536,0.10663821962144639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,51200,0.08442399899164836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,128,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,512,0.0040693332751592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,32,0.0031884445084465873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,51200,0.083533333407508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,12288,0.0273991111252043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,16384,0.01997866729895274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,16384,0.029904001288943823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,10240,0.024157333705160353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,12288,0.01683111157682207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,12288,0.023568888505299885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,8192,0.018912000788582694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,10240,0.020639111598332722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,7168,0.016861332787407767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,8192,0.0177831103404363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,16384,0.034141331911087036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,6144,0.015084443820847405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,7168,0.016007110476493835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,6144,0.01071733319097095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,5120,0.013210666676362356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,10240,0.01519733336236742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,8192,0.01309866706530253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,5120,0.013068444199032254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,4096,0.011735999749766456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,128,0.00543200017677413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,4096,0.008646222452322641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,3584,0.010612444745169746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,6144,0.014477334088749357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,3584,0.008351999852392409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,5120,0.009400000174840292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,51200,0.0510115557246738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,3072,0.009385777844323052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,2560,0.008594666918118795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,3072,0.00797155582242542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,2560,0.007640000018808577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,32,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,2560,0.008559111091825698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,3584,0.009945777555306753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,3072,0.00960000024901496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,2048,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,1536,0.006996444529957241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,1536,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,4096,0.011128000087208219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,768,0.004576000074545543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,1024,0.006664000037643645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,2048,0.007299555672539606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,768,0.006314666734801398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,768,0.00630044440428416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,1536,0.005974222388532426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,256,0.003559111307064692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,512,0.006119110931952794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,512,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,7168,0.01181244436237547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,128,0.0032239999208185407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,1024,0.005274666680230035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,256,0.005561777700980504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,128,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,32,0.0032231110251612137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,1024,0.006365333166387346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,64,0.005558222118351195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,65536,0.08401422368155585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,512,0.004217777815130022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,32,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,2048,0.007336888876226213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,51200,0.06309333112504747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,65536,0.06287555562125312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,65536,0.06810488965776232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,51200,0.049766222635904946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,51200,0.05339822173118591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,16384,0.0199608885579639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,64,0.0032071111102898917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,10240,0.01924711134698656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,12288,0.016515556308958266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,128,0.005456888841258154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,12288,0.016512889001104567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,8192,0.01793244481086731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,10240,0.014833778142929077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,7168,0.01581333412064446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,10240,0.014847111370828418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,8192,0.012459555433856117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,8192,0.013210666676362356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,7168,0.011406222151385413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,16384,0.026532444689008925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,7168,0.01182666669289271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,6144,0.01016444464524587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,256,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,12288,0.022707555029127333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,4096,0.011408000356621213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,5120,0.009355555805895064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,5120,0.00996888925631841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,4096,0.008783111141787635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,4096,0.008714666797055138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,3072,0.009735999835862054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,3584,0.008000888758235509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,6144,0.014367999302016364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,2560,0.00906133320596483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,3072,0.007661333514584436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,5120,0.012780444489585029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,2048,0.007282666862010956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,2560,0.007670222057236566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,16384,0.02032444377740224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,1536,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,3584,0.010028444230556488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,2048,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,1024,0.004906666775544484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,3584,0.008332444561852349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,1024,0.006599999964237213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,1024,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,6144,0.010869333313571082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,768,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,512,0.003906666818592283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,768,0.006342222293217977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,512,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,512,0.005438222239414851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,3072,0.008003555238246918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,256,0.005222222043408289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,128,0.0032071111102898917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,2560,0.0069884442620807225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,2048,0.007334222396214803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,128,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,1536,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,32,0.0032088889016045462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,64,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,768,0.004272888931963178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,65536,0.0747048921055264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,256,0.006017777654859755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,32,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,51200,0.06047022342681885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,65536,0.06127644247478909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,65536,0.0653182201915317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,64,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,16384,0.025654221574465435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,51200,0.048617776897218495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,256,0.003559111307064692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,12288,0.02148355543613434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,16384,0.01980622278319465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,10240,0.01869777838389079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,12288,0.016136000553766888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,10240,0.014164444473054675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,10240,0.014485332700941296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,8192,0.015154666370815702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,128,0.004941333499219683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,8192,0.012418666647540199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,8192,0.012635555532243518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,16384,0.020121778051058452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,6144,0.012420444852775998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,7168,0.011508444117175208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,12288,0.015956444872750174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,5120,0.01110311100880305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,6144,0.010052444206343757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,6144,0.010671111444632212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,4096,0.009669333696365356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,5120,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,3584,0.008994667066468133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,4096,0.008767999708652496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,7168,0.013804444836245643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,3072,0.008051555189821456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,1536,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,3584,0.007666666474607255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,7168,0.011412444214026133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,3072,0.007991111112965478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,3072,0.0074942219588491656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,2048,0.006253333141406377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,2560,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,5120,0.009369778136412302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,1536,0.00526577772365676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,2048,0.007346666521496243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,4096,0.008056888977686564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,51200,0.051296889781951904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,1024,0.004581333448489507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,2560,0.0075146663520071245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,1024,0.006618666566080517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,768,0.0042462220622433555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,1024,0.005665777872006099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,2560,0.007661333514584436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,512,0.00360977773865064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,768,0.006263111200597551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,256,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,512,0.006061333335108227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,2048,0.00656088896923595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,128,0.0031928889867332247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,256,0.005982222242487802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,256,0.0052728888889153796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,64,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,128,0.004943111290534337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,32,0.002874666617976295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,64,0.004937777916590373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,65536,0.06256088945600721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,768,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,32,0.005274666680230035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,512,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,51200,0.04936000042491489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,65536,0.05993600024117363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,1536,0.0069573331210348345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,3584,0.008367111285527548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,51200,0.0477111107773251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,51200,0.04911111129654778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,12288,0.017553778158293832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,16384,0.0192266669538286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,16384,0.01924711134698656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,12288,0.015174221661355762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,12288,0.015743111570676167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,8192,0.015909334023793537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,10240,0.01408622165520986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,65536,0.06223999791675144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,8192,0.01146311147345437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,16384,0.021535999245113794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,128,0.006001777946949005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,6144,0.011736888852384357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,7168,0.010835555692513784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,7168,0.01110577748881446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,6144,0.010049777726332346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,6144,0.010284444524182213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,10240,0.01311466677321328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,4096,0.009023110899660323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,5120,0.009005332986513773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,4096,0.00868888861603207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,7168,0.01531733406914605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,1536,0.006221333311663733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,3584,0.008376888930797577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,3584,0.007640889121426477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,5120,0.010657777388890585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,3072,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,3072,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,8192,0.012074666718641916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,4096,0.008031110796663497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,2560,0.007641777396202087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,2560,0.006924444602595435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,3072,0.007632888853549957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,10240,0.01692800058258904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,1536,0.005245333330498801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,2048,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,5120,0.009384888741705153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,1536,0.0069546666410234236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,1024,0.004442666553788715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,1536,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,768,0.003949333396222857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,2560,0.0069511110583941145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,1024,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,1024,0.005620444400442972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,768,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,768,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,2048,0.006002666635645761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,512,0.006040888941950268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,512,0.0052764444715446895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,128,0.0032275555034478507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,256,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,256,0.004973333328962326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,2048,0.006667555620272954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,128,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,32,0.003212444484233856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,3584,0.008427555362383524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,512,0.003644444462325838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,65536,0.046444445848464966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,32,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,51200,0.03727911247147454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,256,0.0032497776879204642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,16384,0.018053332964579265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,65536,0.059883554776509605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,51200,0.04731200138727824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,12288,0.015211555692884656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,16384,0.018329777651362948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,64,0.003192000091075897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,10240,0.0138586668504609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,12288,0.014524444937705994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,12288,0.015503999259736804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,64,0.004972444640265571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,8192,0.014382221632533602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,10240,0.013456888496875763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,7168,0.01311911145846049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,8192,0.011414222419261932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,65536,0.05816177527109782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,8192,0.011751111182901593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,6144,0.011548444628715515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,51200,0.04663644565476311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,5120,0.010594666832023196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,7168,0.010381333529949188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,6144,0.010120000276300643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,4096,0.009003555609120263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,16384,0.018894222047593858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,5120,0.0092275556590822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,128,0.004943111290534337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,5120,0.009073778159088558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,4096,0.0086666668454806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,3072,0.00740622232357661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,3584,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,3584,0.007638221813572778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,2560,0.006667555620272954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,3072,0.007291555404663086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,7168,0.01076533314254549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,2048,0.005982222242487802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,2560,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,6144,0.009689778089523315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,1536,0.004906666775544484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,3584,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,2048,0.006591111007663939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,1024,0.004583999928500918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,4096,0.008009778128729926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,10240,0.013134222063753339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,1536,0.006112888869312074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,768,0.003909333298603694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,1024,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,3072,0.008030222521887885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,512,0.0035777779089079965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,768,0.006351999938488007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,256,0.0032266666077905228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,512,0.006001777946949005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,2560,0.007671111159854465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,512,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,256,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,2048,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,256,0.005134222408135732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,128,0.005978666659858491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,128,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,1536,0.0069857777820693115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,65536,0.033546666304270424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,64,0.005296000176005893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,32,0.005210666606823604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,1024,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,51200,0.027895109521018133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,65536,0.05752266777886284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,65536,0.05721600188149346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,51200,0.04615199897024366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,128,0.0032186667538351486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,51200,0.04522222280502319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,12288,0.013864888913101621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,16384,0.01685599982738495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,16384,0.018546667363908555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,10240,0.01277511070171992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,12288,0.014171555638313293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,12288,0.014319111903508505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,8192,0.011928000383906893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,10240,0.013062222136391534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,10240,0.01275644451379776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,8192,0.011376000112957425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,8192,0.011043555206722684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,7168,0.010501333408885531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,7168,0.010312000082598792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,64,0.002899555489420891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,32,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,5120,0.009370666411187913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,6144,0.009399111072222391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,5120,0.009364444348547194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,5120,0.00888000014755461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,4096,0.007978666987684038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,7168,0.011387555963463254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,3584,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,4096,0.008667555948098501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,4096,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,3072,0.00666133314371109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,3584,0.007631999750932057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,6144,0.010050666828950247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,3072,0.007349333001507654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,2560,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,2560,0.006863111009200414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,768,0.005579555614127054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,16384,0.01550044450494978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,1536,0.004885333279768626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,2048,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,2048,0.006522666662931442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,3584,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,1536,0.0069573331210348345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,1536,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,1024,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,1024,0.005618666609128316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,768,0.006264888991912206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,768,0.005355555564165115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,512,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,6144,0.010448888772063784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,256,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,512,0.006299555715587404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,2560,0.005986666513813867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,256,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,256,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,2048,0.005296000176005893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,128,0.004959110998445087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,64,0.0029093333416514923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,32,0.0029111111329661477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,64,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,32,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,3072,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,128,0.002867555452717675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,768,51200,0.02558311157756382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,65536,0.057143999470604785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,512,0.0036346668170558084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,768,16384,0.01329066687160068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,51200,0.04442755712403191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,16384,0.016788444585270353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,16384,0.01721955504682329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,768,65536,0.030906667311986286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,768,0.003908444609906939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,12288,0.014365333649847241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,12288,0.014143110977278816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,65536,0.055811557504865855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,768,8192,0.011386666860845355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,10240,0.01277955538696713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,10240,0.012429333395428128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,768,7168,0.011040888726711273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,8192,0.011384888655609555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,8192,0.01109955542617374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,1024,0.004238222208287981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,768,12288,0.011591110792424945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,7168,0.010220444036854638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,768,5120,0.008944888909657797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,6144,0.009827555881606208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,6144,0.009382222261693742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,768,4096,0.007640889121426477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,5120,0.00868888861603207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,128,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,768,3584,0.007303111255168915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,4096,0.008671110702885522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,3584,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,7168,0.01070666644308302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,768,10240,0.01313866674900055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,5120,0.009038222332795462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,768,2560,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,51200,0.045886221859190196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,4096,0.00796977761718962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,768,2048,0.006581333362393909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,2560,0.006879999819729064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,768,6144,0.010031999813185798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,768,1536,0.005636444522274866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,2048,0.006346666564544042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,3072,0.007256888680987888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,1536,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,1536,0.006959999601046245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,768,1024,0.004583999928500918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,1024,0.00664533343580034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,768,768,0.0038968887594011095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,768,3072,0.006660444455014334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,2048,0.0070222218831380205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,768,512,0.003528000166018804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,768,0.005618666609128316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,512,0.006010666489601135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,768,256,0.003109333415826162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,256,0.005247999810510212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,3584,0.007522666619883643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,256,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,768,0.006268444574541516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,128,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,768,64,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,512,0.0052826665341854095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,768,32,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,512,65536,0.0251555558707979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,64,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,2560,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,32,0.004976888911591636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,65536,0.057146668434143066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,65536,0.054393778244654335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,512,16384,0.013793778088357715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,51200,0.04388088981310526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,128,0.004937777916590373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,16384,0.016891555653678048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,16384,0.016857778032620747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,12288,0.014141332772043018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,12288,0.014164444473054675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,512,51200,0.020273778173658583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,768,128,0.0028444443725877335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,10240,0.013006221916940479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,10240,0.012736000120639801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,8192,0.011416888899273343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,512,12288,0.012186666329701742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,512,7168,0.00904622260067198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,7168,0.01074577785200543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,512,6144,0.009069333473841349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,512,8192,0.009702222214804756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,6144,0.009467555416954888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,51200,0.04366222355100843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,512,5120,0.00868266655339135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,6144,0.00998488896422916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,512,4096,0.007864000068770515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,5120,0.009346666435400644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,8192,0.01072622256146537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,5120,0.008671999805503422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,512,3584,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,4096,0.00870222184393141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,4096,0.007984889050324758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,3584,0.00833155545923445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,3584,0.007640000018808577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,512,2560,0.005858666780922148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,3072,0.00720888872941335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,2560,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,2560,0.006728000111050076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,3072,0.008016889293988546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,2048,0.007336888876226213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,2048,0.006404444575309753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,7168,0.010058666268984476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,1536,0.0069448889957533936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,512,1024,0.0039022221333450745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,512,2048,0.005210666606823604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,1024,0.005619555711746216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,512,768,0.003555555724435382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,1024,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,512,3072,0.006303999986913469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,768,0.005573333137565189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,3072,0.008000888758235509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,512,0.0052728888889153796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,512,256,0.003173333282272021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,512,0.006242666807439592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,256,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,512,128,0.0028693332440323303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,512,1536,0.004566222015354368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,512,512,0.0032462223122517266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,512,64,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,128,0.00602311102880372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,1536,0.005977777971161737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,64,0.005120888766315249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,256,65536,0.020640888147883944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,32,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,256,0.005977777971161737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,256,51200,0.01787911189926995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,65536,0.05382400088840061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,512,10240,0.010724444356229572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,256,16384,0.011746666497654386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,51200,0.040080888403786555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,128,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,51200,0.04315111041069031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,16384,0.016912889149453905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,16384,0.016553777787420485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,256,10240,0.009369778136412302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,12288,0.014206222361988492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,12288,0.01386666629049513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,10240,0.01312800000111262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,65536,0.04969333277808296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,10240,0.01240533341964086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,768,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,256,7168,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,8192,0.010792888700962067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,7168,0.01071200023094813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,512,32,0.0029048888633648553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,7168,0.010040000081062317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,256,5120,0.008039111064540016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,6144,0.010095111197895473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,6144,0.009356444080670675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,256,4096,0.007584888901975419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,5120,0.009367999931176504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,5120,0.00869422240389718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,256,8192,0.00850844466023975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,4096,0.008719999757077958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,4096,0.007753777835104201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,8192,0.01111822244193819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,256,3072,0.006265777680608962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,3584,0.007671111159854465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,256,6144,0.007872888611422645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,256,2560,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,3072,0.0069822221994400024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,2560,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,2560,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,2048,0.007318221860461765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,2048,0.006643555644485686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,256,1536,0.004551110996140374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,1536,0.0069466663731469055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,3584,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,256,1024,0.0038897775941424896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,256,768,0.00360000009338061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,256,12288,0.010062221851613786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,1024,0.005676444619894028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,768,0.006324444380071428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,256,512,0.0033288889875014624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,768,0.0053591111467944244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,256,256,0.002862222285734283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,512,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,512,0.005294222384691238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,256,128,0.0029599999801980127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,256,0.004956444518433677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,1536,0.006016888966162999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,256,64,0.002569777684079276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,3072,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,256,32,0.002579555536309878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,256,2048,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,64,0.004969777746333016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,32,0.004979555391603046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,256,3584,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,256,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,65536,0.051648888323042125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,51200,0.03980888923009237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,128,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,1024,0.006647999915811751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,16384,0.016920889417330425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,128,65536,0.014846222268210517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,128,51200,0.01311555587583118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,12288,0.014151111245155334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,128,0.005921777751710679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,128,10240,0.009355555805895064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,10240,0.012789333032237159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,10240,0.012440000143316058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,16384,0.01686311099264357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,8192,0.01130844404300054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,8192,0.010747555229398938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,128,7168,0.007714666426181793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,12288,0.013518222504191928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,128,6144,0.008678221868144141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,7168,0.01074577785200543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,7168,0.010040889183680216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,128,16384,0.009523555636405945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,6144,0.010006222460005019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,6144,0.009005332986513773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,128,4096,0.007303111255168915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,5120,0.009353777600659264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,5120,0.008672888908121321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,128,3584,0.006960888703664144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,4096,0.007663999994595845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,65536,0.049327999353408813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,3584,0.008026666939258575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,51200,0.041406220859951444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,128,5120,0.007970666719807519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,3072,0.0069351109365622205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,3072,0.007889777421951294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,128,8192,0.008356444537639618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,128,2048,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,2560,0.007637333538797166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,4096,0.00868711123863856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,128,1536,0.004223111189073986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,2048,0.007268444531493717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,2048,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,128,1024,0.003882666842805015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,1536,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,1024,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,1024,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,128,12288,0.009757333331637913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,2560,0.006538666784763336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,768,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,768,0.006326222171386083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,128,3072,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,512,0.0052933332820733385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,512,0.005987555616431766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,128,2560,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,256,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,128,128,0.002570666579736604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,128,512,0.0032248888164758682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,128,0.005671999934646819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,128,64,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,128,256,0.002903999967707528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,128,32,0.002564444517095884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,128,768,0.0035413333939181436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,64,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,256,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,64,51200,0.014109333356221518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,65536,0.05064355664783054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,64,16384,0.009349333743254343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,51200,0.041060444381501936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,64,12288,0.009703111317422655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,64,10240,0.009009777671760982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,16384,0.01651822196112739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,64,8192,0.007979555262459649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,12288,0.013601777454217276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,64,65536,0.014927110738224454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,64,7168,0.008971555365456475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,32,0.004971555537647671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,64,6144,0.008365333080291748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,8192,0.010489777558379702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,64,5120,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,64,4096,0.007320889168315464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,128,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,6144,0.009359999663299983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,64,3584,0.006955555743641323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,64,3072,0.006753777878151999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,4096,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,64,2560,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,10240,0.012362666428089142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,64,2048,0.005032000028424793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,2560,0.006433777925040986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,64,1536,0.004579555657174853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,7168,0.010040000081062317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,2048,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,3584,0.007353777686754863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,64,768,0.003579555700222651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,1536,0.0057946667075157166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,5120,0.00832177781396442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,64,512,0.0032017777363459268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,64,256,0.002918222298224767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,3584,0.007384888827800751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,64,128,0.0029022221763928733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,512,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,64,64,0.002502222235004107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,256,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,64,32,0.002616000051299731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,128,0.004941333499219683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,64,1024,0.003607999947335985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,1536,0.00692622239391009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,32,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,32,51200,0.014128888646761576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,32,16384,0.009344889058007134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,1024,0.005625777774386936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,65536,0.050717333952585854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,32,12288,0.009675555759006077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,32,10240,0.00889066689544254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,3072,0.006979555719428592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,16384,0.016563556260532804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,12288,0.013818666338920593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,32,7168,0.008966222405433655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,10240,0.012362666428089142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,32,6144,0.0083137775460879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,8192,0.01070666644308302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,32,5120,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,7168,0.010001777774757808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,32,4096,0.007098666495747036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,5120,0.008469333251317343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,51200,0.04076711005634732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,4096,0.007748444047239091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,32,3072,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,3584,0.007324444750944774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,32,2560,0.005585777676767773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,3072,0.007159111400445302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,64,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,32,2048,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,32,1536,0.004569777597983678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,6144,0.009375111096435124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,32,1024,0.0036888888312710654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,2048,0.006137777947717243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,32,768,0.0035386664999855887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,32,3584,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,32,512,0.003195555466744635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,32,8192,0.00798044436507755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,32,256,0.002966222249799304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,768,0.00526577772365676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,32,128,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,512,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,32,64,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,256,0.00499911109606425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,32,32,0.002508444504605399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,128,0.004819555415047539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,2560,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,64,0.004912000149488449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,1536,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,768,0.005550222264395819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,1024,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,161,32,65536,0.015192889504962497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,51200,1.0847991307576497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,51200,0.8481911023457845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,16384,0.27408088578118217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,16384,0.13685066170162624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,51200,0.4061804347568088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,12288,0.2658906777699788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,16384,0.3563235600789388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,12288,0.20662666691674125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,10240,0.23516978157891166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,12288,0.10513866609997219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,32,0.004632888982693355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,8192,0.19514666663275826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,10240,0.1715164449479845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,16384,0.31283820999993217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,12288,0.2290426625145806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,7168,0.167449778980679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,10240,0.2002035511864556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,7168,0.12446133295694987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,8192,0.155039111773173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,10240,0.08857955535252889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,7168,0.06466133064693876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,6144,0.05638844437069363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,6144,0.10644355747434829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,8192,0.13884533776177302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,5120,0.12716621822781032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,6144,0.12088889545864528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,4096,0.10276444753011067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,6144,0.14361688825819227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,5120,0.09927822483910455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,4096,0.03983911209636264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,4096,0.07115733623504639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,7168,0.14050489001803929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,3584,0.08725155724419488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,4096,0.08142755428949992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,3072,0.07665599717034234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,3584,0.06440444125069512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,3584,0.07351822323269315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,3072,0.05460355679194132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,2560,0.06290666924582587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,3072,0.06451111369662814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,8192,0.07128089004092746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,2560,0.02784444557295905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,2560,0.04757600029309591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,2560,0.05525066786342197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,3584,0.035417778624428645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,2048,0.023912888434198167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,1536,0.040002667241626315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,2048,0.046952890025244824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,1536,0.03199288911289639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,1536,0.03917600048912896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,5120,0.047798222965664335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,1024,0.030403554439544678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,51200,0.9192960527208117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,2048,0.04051911168628269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,768,0.022608000371191237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,1024,0.03162489003605313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,3072,0.031558222240871854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,768,0.014200000299347771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,768,0.025106666816605464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,768,0.02843733297453986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,5120,0.08959466881222194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,2048,0.052375111314985484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,256,0.010407110883129967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,512,0.013209777573744455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,512,0.02348355617788103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,256,0.01572355628013611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,128,0.009284444153308868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,256,0.01909955508179135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,512,0.01651022169325087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,128,0.014817777607176038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,64,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,1536,0.0188426673412323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,512,0.017870222528775532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,32,0.00849599970711602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,64,0.014816888504558139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,32,0.015440889530711703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,256,0.011786667009194693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,128,0.01646133263905843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,65536,1.148249732123481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,65536,0.853691577911377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,1024,0.024933333198229473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,51200,0.9009528689914279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,65536,0.41443644629584414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,128,0.01179377817445331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,16384,0.2917528947194417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,1024,0.015722667177518208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,16384,0.2274035612742106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,12288,0.2217359940210978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,16384,0.10883200168609619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,12288,0.17068621847364637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,51200,0.6686151292588977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,51200,0.3303617901272244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,16384,0.253166225221422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,10240,0.13438577122158474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,12288,0.18570221794976127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,10240,0.06873422198825412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,8192,0.14041954941219753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,12288,0.08194044563505384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,10240,0.16173156102498373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,8192,0.11355911360846625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,7168,0.1279484430948893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,7168,0.09585333532757229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,8192,0.12552711698744032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,10240,0.17028444343143037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,7168,0.050343109501732715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,7168,0.11542133490244548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,8192,0.056576000319586865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,6144,0.044974221123589404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,51200,0.7582809130350748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,5120,0.09329155418607925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,65536,0.9778436024983724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,6144,0.09568533632490371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,5120,0.037976887491014265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,6144,0.10587644577026367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,4096,0.03274755676587423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,4096,0.06849066416422527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,6144,0.08304178052478366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,3584,0.052087999052471586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,3584,0.029289778735902574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,4096,0.07678577635023329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,3584,0.06150044335259331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,3072,0.04539022180769178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,2560,0.049570666419135205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,3584,0.0651351081000434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,3072,0.05497422152095371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,2560,0.038945777548684016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,2048,0.040580444865756564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,2560,0.04882399903403389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,2048,0.033450666401121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,3072,0.026280000805854797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,2048,0.04186488853560554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,5120,0.08237955305311415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,2048,0.019870221614837646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,1536,0.01648622254530589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,2560,0.023761777414215937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,1024,0.024886222349272832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,1536,0.03439377744992574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,1024,0.013787555197874704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,1024,0.020717332760492962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,768,0.01847644481394026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,1024,0.02764355474048191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,1536,0.032009777095582746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,768,0.012803555362754397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,768,0.024255999260478552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,1536,0.026882666680547927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,512,0.015263999501864115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,3072,0.057690666781531445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,512,0.011775111158688864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,512,0.02050488856103685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,256,0.010436444646782346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,256,0.013103110922707452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,128,0.007762666377756331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,256,0.017172444197866652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,4096,0.057943112320370145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,512,0.012844444149070315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,128,0.010659555594126383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,128,0.01240799989965227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,32,0.006265777680608962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,128,0.015009777413474189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,64,0.0124195557501581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,5120,0.0739484429359436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,32,0.013367999758985309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,65536,0.380038234922621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,65536,0.30315910445319283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,51200,0.29706843694051105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,65536,0.14350933498806423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,64,0.005979555762476391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,16384,0.10306577550040351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,51200,0.23448710971408418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,51200,0.11305067274305557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,256,0.009380444056457942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,16384,0.08704533179601033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,12288,0.060144000583224826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,768,0.020632889535692003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,12288,0.06701778040991889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,10240,0.06311200062433879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,16384,0.04012444284227159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,10240,0.05751022365358141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,10240,0.027898665931489732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,8192,0.052863111098607383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,8192,0.04361244373851352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,65536,0.32189154624938965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,51200,0.25511643621656627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,8192,0.04876177840762668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,8192,0.023579556081030104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,10240,0.05120533373620775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,7168,0.02130577796035343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,6144,0.04105688797103034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,7168,0.04361866580115425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,6144,0.018983999888102215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,5120,0.034727109803093806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,6144,0.03926488757133484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,12288,0.03218044506178962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,7168,0.047169778082105845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,5120,0.028872887293497723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,5120,0.016903110676341586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,4096,0.02878755662176344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,7168,0.03779822256830003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,4096,0.015156444576051501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,4096,0.024375999967257183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,3584,0.026027555267016094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,6144,0.033533334732055664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,3584,0.013786666923099093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,3584,0.022056889202859666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,3072,0.022766222556432087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,3584,0.026867555247412786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,5120,0.034444444709353976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,3072,0.01946311030122969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,2560,0.01925066610177358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,3072,0.02500177754296197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,2560,0.012059555285506778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,2560,0.017064889272054035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,2048,0.0159608887301551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,2048,0.014879110786649915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,2048,0.020080000162124634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,2048,0.011132444772455426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,1536,0.012709333664841123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,4096,0.029824889368481104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,2560,0.02233155568440755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,1024,0.01036088913679123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,1536,0.009022222624884712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,1536,0.017258667283587985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,1024,0.010693333215183683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,768,0.008649778034951951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,1024,0.0148062225845125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,1536,0.013110222087966071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,512,0.0069431112044387394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,768,0.007709333466158972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,768,0.009689778089523315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,768,0.014350222216712104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,256,0.005158222383923001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,512,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,512,0.012131555212868584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,256,0.007471111085679796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,1024,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,12288,0.07767733600404528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,128,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,128,0.007171555525726742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,512,0.007300444775157505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,64,0.0038933331767717996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,128,0.009696000152164036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,32,0.0038942222793896994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,64,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,128,0.004260444392760594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,16384,0.07930399974187215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,65536,0.22119555208418104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,65536,0.12595289283328587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,256,0.006984000404675801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,256,0.010377777947319878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,51200,0.1704168849521213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,3072,0.013064000341627332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,65536,0.30885688463846844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,16384,0.05818311373392741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,51200,0.21513956122928193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,12288,0.057321780257754855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,16384,0.037584000163608126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,16384,0.08355466524759929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,32,0.007803555164072249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,12288,0.03006311257680257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,12288,0.044093334012561373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,12288,0.06532533301247491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,51200,0.10105422470304702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,8192,0.04052177733845181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,10240,0.05615288681454129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,8192,0.032848887973361544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,65536,0.27535467677646214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,8192,0.04758222235573662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,8192,0.021888888544506494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,7168,0.0351839992735121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,7168,0.01993333299954732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,6144,0.030755556292004053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,10240,0.038000000847710505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,6144,0.02528800070285797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,6144,0.03788444399833679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,51200,0.24062132835388184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,6144,0.018171555466122098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,10240,0.047468443711598717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,16384,0.07246933380762736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,4096,0.022636445032225713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,5120,0.016194666425387066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,5120,0.021524444222450256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,5120,0.033313777711656355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,10240,0.02552000019285414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,4096,0.014634667171372307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,4096,0.018411555223994784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,4096,0.02870044443342421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,3072,0.017542221479945712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,3584,0.016559999850061204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,3584,0.026632888449562922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,5120,0.026509332987997267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,2560,0.014839111102951897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,3072,0.012673777838548025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,3072,0.023989333046807185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,7168,0.028520888752407495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,2560,0.013568888935777875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,2560,0.011390222443474663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,2560,0.021701332595613267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,3584,0.013449778159459433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,2048,0.012087110843923358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,1536,0.010710222025712332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,2048,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,3072,0.015220445063379077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,1536,0.010535111029942831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,1536,0.017258667283587985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,7168,0.04316622349951002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,2048,0.012791111237472959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,1024,0.008661333057615492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,768,0.007700444095664554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,1024,0.013817778064144982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,2048,0.019632889164818656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,768,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,512,0.006164444403515921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,768,0.01347733371787601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,1024,0.008988444175985124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,3584,0.01959733333852556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,256,0.004571555389298333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,512,0.007290666302045186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,512,0.011761777930789523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,256,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,256,0.010096888575288985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,768,0.007661333514584436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,128,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,1536,0.008465777668688033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,64,0.003903111235962974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,128,0.009678222239017487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,32,0.004059555629889171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,512,0.007300444775157505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,64,0.006659555352396435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,65536,0.24815734227498373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,256,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,128,0.0038862224254343244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,51200,0.19342488712734648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,128,0.00702044450574451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,51200,0.1602835522757636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,16384,0.06496089034610324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,1024,0.007706666986147563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,65536,0.30510934193929035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,16384,0.05515999926461113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,65536,0.20544800493452284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,12288,0.05137688914934794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,16384,0.03632355398601956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,51200,0.2389733261532254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,16384,0.08326399988598294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,10240,0.04256355431344774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,12288,0.042336000336541064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,12288,0.02836622132195367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,8192,0.036534223291609026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,10240,0.03646311163902283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,32,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,8192,0.031143112315071955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,8192,0.0215013325214386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,8192,0.0469742218653361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,65536,0.12395377953847249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,7168,0.02699377801683214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,7168,0.019371555911170114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,7168,0.041894220643573336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,12288,0.06458133459091187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,10240,0.024837333295080397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,5120,0.02402488887310028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,6144,0.023804444405767653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,6144,0.03716355562210083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,7168,0.031947556469175555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,5120,0.020541333489947848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,4096,0.020613332589467365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,5120,0.015995555453830294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,4096,0.014362666341993542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,4096,0.017644445101420086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,3584,0.017960889471901786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,4096,0.027965333726671007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,3584,0.013117333253224691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,3584,0.016190222567982145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,3584,0.0258933338854048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,3072,0.014856888188256158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,5120,0.03278933299912347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,2560,0.01368533323208491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,3072,0.02342133389578925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,3072,0.012074666718641916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,6144,0.027426666683620874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,2048,0.012406222522258759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,2560,0.013384888569513956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,2560,0.02115733259254032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,6144,0.01754755609565311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,2048,0.011779555843936073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,2048,0.009568000005351173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,2048,0.018928888771269057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,1536,0.010033778018421596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,1024,0.008634666601816813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,1536,0.016563556260532804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,2560,0.01071466671095954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,1024,0.008231999973456064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,768,0.006607999818192587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,1024,0.007689778175618913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,51200,0.0976782242457072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,1536,0.010210666391584609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,768,0.007820444802443186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,768,0.007332444190979004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,512,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,768,0.012332444389661154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,1536,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,256,0.004567111117972268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,512,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,512,0.006955555743641323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,512,0.011037333144081963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,128,0.00397866674595409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,256,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,256,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,256,0.00976533359951443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,10240,0.05619555711746216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,128,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,128,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,32,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,128,0.009340444372759925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,65536,0.18635733922322592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,32,0.006710222197903528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,65536,0.16364888350168863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,65536,0.07628799809349908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,1024,0.012993777791659037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,3072,0.015820443630218506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,51200,0.12694756189982095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,65536,0.16558755768669978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,64,0.00664444433318244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,16384,0.044525331921047635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,51200,0.13220800293816462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,12288,0.03894044293297662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,16384,0.04793333344989353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,64,0.003974222060706881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,12288,0.019540444016456604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,12288,0.03468888998031616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,12288,0.03757777810096741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,16384,0.05050666795836555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,8192,0.028174221515655518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,10240,0.03296355406443278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,16384,0.023707555400000677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,8192,0.024991111622916326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,7168,0.02460977766248915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,8192,0.014977777997652689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,10240,0.032771554258134626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,7168,0.021820444199774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,6144,0.02129688858985901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,7168,0.026149332523345947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,10240,0.01721244388156467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,6144,0.012813333008024426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,5120,0.018358222312397428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,6144,0.02384888960255517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,8192,0.028768890433841284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,5120,0.01717511150572035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,4096,0.015691555208630033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,5120,0.011905777785513135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,7168,0.013834666874673633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,10240,0.029715554581748113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,4096,0.014903111590279473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,3584,0.01384888920519087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,4096,0.019235556324323017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,6144,0.01953066719902886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,3072,0.012499555945396423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,51200,0.14234933588239881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,3584,0.013830222189426422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,3584,0.01737777723206414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,2560,0.010703110860453712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,3072,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,5120,0.021549334128697712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,3072,0.016174222032229107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,51200,0.06071911255518595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,2560,0.011411555939250521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,2560,0.008303110798199972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,2560,0.014828445182906257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,4096,0.011015999648306104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,1536,0.008163555628723567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,2048,0.007687110867765214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,2048,0.010393777655230628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,1024,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,1536,0.008652444514963362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,1536,0.012463111016485425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,3072,0.012753778033786349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,1024,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,1024,0.010217777556843227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,2048,0.009407111340098912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,768,0.0073315559162033936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,768,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,768,0.009703111317422655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,2048,0.013844444519943662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,512,0.0064488889442549805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,256,0.003875555677546395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,512,0.006392000036107169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,512,0.009352888498041365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,768,0.005241777747869492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,128,0.0035413333939181436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,256,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,256,0.005992888990375731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,256,0.008656000097592672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,64,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,512,0.00425155543618732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,32,0.0032195556494924757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,128,0.008307555483447181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,3584,0.009920888476901585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,32,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,1024,0.0069271110826068455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,65536,0.15612267123328313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,128,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,128,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,51200,0.13308710522121853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,65536,0.07177778085072835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,1536,0.007365333537260692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,16384,0.04528533418973287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,65536,0.15350577566358778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,51200,0.05822933382458157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,16384,0.022776888476477727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,51200,0.11915466520521377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,12288,0.03647111190689935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,16384,0.04544711112976074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,12288,0.01880088945229848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,10240,0.030371556679407757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,64,0.006309333360857434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,51200,0.12307555145687527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,65536,0.16886844899919298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,8192,0.026379555463790894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,10240,0.028538667493396338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,10240,0.03295822276009454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,16384,0.04305511050754123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,8192,0.014749333262443542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,8192,0.028227554427252874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,7168,0.021292444732454088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,7168,0.025461332665549383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,7168,0.013636444177892474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,6144,0.019186667270130582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,10240,0.016525333126386006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,6144,0.023568888505299885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,6144,0.012611555556456247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,8192,0.024392889605628118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,5120,0.011490667031870948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,5120,0.01685422162214915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,4096,0.015200000670221118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,12288,0.0333591103553772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,5120,0.021159110797776118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,6144,0.01998133295112186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,3584,0.012799111505349478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,4096,0.010026666853162978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,4096,0.014782221780882942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,4096,0.01863822175396813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,3584,0.009352000223265754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,3584,0.01725955473052131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,5120,0.01736088924937778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,2560,0.010442666709423065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,3072,0.012505778008037143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,3072,0.008786666724416945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,3072,0.016011555989583332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,2560,0.0147386673423979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,12288,0.03718400001525879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,3072,0.012109333442317115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,3584,0.013550221920013428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,1536,0.007968000239796108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,2048,0.013445333474212222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,2048,0.007686222592989604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,7168,0.022825777530670166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,1024,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,1536,0.0073626670572492815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,1536,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,1536,0.011770666473441653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,768,0.004912000149488449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,1024,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,2560,0.008006222546100616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,1024,0.010065777434243096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,512,0.0042568888101312844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,768,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,2048,0.010037333601050908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,512,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,512,0.006653333289755716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,256,0.003896000070704354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,512,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,256,0.005993777679072485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,256,0.005972444597217772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,2048,0.009299555586444007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,768,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,128,0.005995555470387141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,768,0.010049777726332346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,128,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,128,0.008051555189821456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,32,0.003567111161020067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,64,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,32,0.006292444550328785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,2560,0.011400889191362592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,256,0.008012444608741337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,65536,0.11858844757080078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,51200,0.12160000536176895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,65536,0.06934577888912626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,64,0.003569777641031477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,16384,0.042415999703937106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,51200,0.056438220871819385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,65536,0.15316355228424072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,51200,0.11303022172715928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,16384,0.033724443780051336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,12288,0.033744888173209295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,16384,0.04350489046838549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,12288,0.025937777426507738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,12288,0.034669333034091525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,10240,0.028203556934992473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,12288,0.018192888961897958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,10240,0.02252888845072852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,16384,0.022091555926534865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,10240,0.01605422298113505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,8192,0.02441066669093238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,8192,0.019376888871192932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,8192,0.0262044436401791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,51200,0.09241066376368205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,8192,0.014271999398867289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,7168,0.016866667403115165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,1024,0.006656888872385025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,7168,0.02345955537425147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,6144,0.018584888842370775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,7168,0.013292444248994192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,6144,0.012155555188655853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,5120,0.01605599953068627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,6144,0.02182933357026842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,65536,0.14197066095140246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,4096,0.013961777091026306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,5120,0.011114666859308878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,5120,0.01382844481203291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,5120,0.01960711181163788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,3584,0.012127999630239276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,128,0.003871110992299186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,4096,0.01761777698993683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,3584,0.009369778136412302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,6144,0.015503999259736804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,3584,0.016169778174824186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,7168,0.02136444383197361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,3072,0.0107360002067354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,3072,0.008346666892369589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,2560,0.011398221883508893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,4096,0.012436444560686747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,10240,0.030247999562157526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,2048,0.010107555323176915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,2560,0.007640889121426477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,3584,0.011722666521867117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,3072,0.011416888899273343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,1536,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,2048,0.00739555557568868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,2048,0.012447111308574677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,1536,0.006952889263629913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,1024,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,4096,0.009752888646390703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,2560,0.009470221896966299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,2560,0.01404355631934272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,1024,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,1024,0.006677333265542984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,1024,0.009709333380063375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,2048,0.008223999705579545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,512,0.004902222090297275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,768,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,768,0.009333333207501305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,3072,0.015000889698664347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,1536,0.010718222293588849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,512,0.00868888861603207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,256,0.005592888842026393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,256,0.007640889121426477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,256,0.005980444451173146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,128,0.0038817777401871155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,768,0.006474666711356904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,128,0.005583999885453119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,1536,0.007334222396214803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,64,0.0033244445092148254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,512,0.005990222096443176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,32,0.003228444399105178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,65536,0.14248889022403294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,768,0.005617777920431561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,65536,0.11530044343736436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,128,0.005964444329341252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,65536,0.06709777646594577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,512,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,51200,0.1137617826461792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,64,0.005653333332803514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,51200,0.09073688586552937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,51200,0.05458577805095249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,51200,0.1117982202106052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,16384,0.03257600135273404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,16384,0.021129777034123737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,16384,0.04321777820587158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,256,0.004202666795916027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,12288,0.025032889511850145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,12288,0.017492444978819955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,12288,0.03421777817938063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,32,0.005984000033802456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,10240,0.015863110621770222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,8192,0.0255422227912479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,12288,0.036425779263178505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,65536,0.14027644528283015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,8192,0.025849777791235182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,7168,0.022568888134426538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,16384,0.04569866591029697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,10240,0.021544888615608215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,7168,0.016976000534163583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,10240,0.030066665675905015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,6144,0.019680000013775296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,7168,0.023220444718996685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,128,0.00795555528667238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,6144,0.021336888273557026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,8192,0.013953778478834363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,5120,0.013539555172125498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,7168,0.012737777498033313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,5120,0.010415111151006486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,8192,0.018704000446531508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,5120,0.016902221573723685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,4096,0.012167111039161682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,4096,0.01754844519827101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,3584,0.012448000411192576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,5120,0.019136000010702345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,6144,0.01532977819442749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,3072,0.01144444445768992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,3584,0.008415111237102086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,3584,0.016176889340082806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,3072,0.008216000265545314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,2560,0.010052444206343757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,3072,0.01482222146458096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,4096,0.009394666386975182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,2560,0.007879111501905654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,2048,0.008745777938101027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,4096,0.014363555444611443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,10240,0.03028977910677592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,2048,0.008175111479229396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,1536,0.007431999676757389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,2048,0.012469333079126147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,3072,0.010289777484205034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,1536,0.006940444310506185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,1536,0.007344000041484833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,1024,0.005759999983840519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,1536,0.010718222293588849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,2560,0.013526221944226159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,1024,0.006601777755551868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,1024,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,1024,0.009721777505344814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,6144,0.012074666718641916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,512,0.004263111286693149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,768,0.0063973334100511335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,2048,0.007332444190979004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,768,0.009053332938088311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,256,0.0038942222793896994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,512,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,512,0.00849955528974533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,2560,0.009696000152164036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,128,0.003571555432346132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,768,0.004945777770545748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,128,0.005586666779385672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,768,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,128,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,64,0.0032115555885765287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,32,0.0032151111712058387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,512,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,64,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,3584,0.011412444214026133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,65536,0.10942311419381036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,128,0.006005333529578314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,65536,0.14062310589684382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,65536,0.06442222330305311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,32,0.0063164445261160536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,51200,0.08555022213194106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,256,0.007669332954618666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,16384,0.03419377737575107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,51200,0.11091644234127468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,16384,0.030933333767784968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,16384,0.02036888897418976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,12288,0.027642667293548584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,51200,0.08065866761737399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,12288,0.023766222927305434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,12288,0.01717422240310245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,10240,0.02018311123053233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,256,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,10240,0.015321777926550971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,10240,0.02106933295726776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,8192,0.01713244451416863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,10240,0.029270221789677937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,8192,0.013170666992664337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,8192,0.018218666315078735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,8192,0.02539644473128849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,12288,0.03359999921586778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,256,0.005642666584915585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,6144,0.013746666411558786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,7168,0.012114666402339935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,51200,0.051889777183532715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,6144,0.014852444330851236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,5120,0.01180355581972334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,7168,0.01514933341079288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,5120,0.013457777599493662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,7168,0.016223111086421542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,5120,0.019097778532240126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,7168,0.023003555006451074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,4096,0.011997333003415002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,6144,0.011055111057228513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,4096,0.016977777083714802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,3584,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,3584,0.01052266690466139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,3584,0.015503999259736804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,65536,0.10366221931245591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,3584,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,5120,0.00962755580743154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,3072,0.009723555710580613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,3072,0.014477334088749357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,16384,0.04194577866130405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,3072,0.008097777764002482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,2560,0.008658666577604082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,2560,0.01313866674900055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,3072,0.00886222223440806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,2560,0.007694222033023834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,2048,0.008026666939258575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,2048,0.01149866647190518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,6144,0.020899555749363367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,2048,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,1536,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,1536,0.010758221977286868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,2048,0.006614222294754452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,1024,0.00479644454187817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,1024,0.006590222318967183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,1024,0.009672889278994666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,2560,0.007962666451931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,768,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,4096,0.010423111418883005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,768,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,768,0.009111111362775167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,512,0.005967999911970562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,768,0.0038968887594011095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,512,0.008661333057615492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,256,0.0035546666218174826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,256,0.005635555419656966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,256,0.00795377790927887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,1536,0.0069848886794514125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,256,0.005994666781690385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,128,0.005446222093370225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,4096,0.00904622260067198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,128,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,64,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,32,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,512,0.0038542221817705366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,65536,0.09228889147440593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,32,0.005668444352017508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,1536,0.005695999910434087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,512,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,65536,0.10623821947309707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,51200,0.07535911268658109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,65536,0.06352888875537448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,65536,0.14015289147694907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,16384,0.03153866529464722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,51200,0.08345333072874282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,128,0.0032071111102898917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,51200,0.11077688799964057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,16384,0.020392888122134738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,16384,0.029784000582165186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,16384,0.042111999458736844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,1024,0.00665866666369968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,64,0.002874666617976295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,10240,0.022221333450741235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,12288,0.016901334126790363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,12288,0.03341155582004123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,8192,0.015587555037604438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,10240,0.015209777487648858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,10240,0.020448888341585796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,51200,0.05093333456251356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,7168,0.014409777190950183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,8192,0.01778133379088508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,12288,0.024703999360402424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,8192,0.02545333405335744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,7168,0.016120889120631747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,128,0.008035555481910706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,12288,0.023473777704768713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,6144,0.014731556177139282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,5120,0.011603555745548673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,6144,0.020991999242040846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,10240,0.0291608903143141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,8192,0.01293155550956726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,5120,0.013153778182135688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,5120,0.019193778435389202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,6144,0.012704888979593912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,7168,0.022833777798546687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,3584,0.010828444527255164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,6144,0.01071999967098236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,4096,0.017256889078352187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,3584,0.010455999937322406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,3584,0.008437333007653555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,3584,0.015125332607163323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,3072,0.009374221993817223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,7168,0.012090666426552666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,2560,0.007284444239404466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,3072,0.01390044391155243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,4096,0.008719111482302347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,2560,0.007690666450394525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,2560,0.008437333007653555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,2048,0.006383111079533895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,2560,0.012846222354306115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,3072,0.008654221892356873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,1536,0.00552977787123786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,2048,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,2048,0.007815111014578078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,2048,0.011758222348160215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,1024,0.004338666796684265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,1536,0.0069822221994400024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,4096,0.011050666371981302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,1536,0.010568000376224518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,3072,0.0081404447555542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,768,0.003912888881233004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,1024,0.00656088896923595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,1024,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,1024,0.009400000174840292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,512,0.0037128888070583344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,768,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,768,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,768,0.009062222308582729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,4096,0.01184266640080346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,512,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,512,0.006031111296680238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,512,0.008657777474986183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,1536,0.00721066693464915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,128,0.0032222221295038858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,256,0.00543200017677413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,256,0.00776622196038564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,64,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,128,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,128,0.0052897776994440295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,32,0.0032035555276605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,128,0.008003555238246918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,65536,0.07787822352515326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,64,0.00544799988468488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,32,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,5120,0.009393778112199571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,51200,0.0646373364660475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,65536,0.06258044640223186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,256,0.006016888966162999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,51200,0.050195554892222084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,51200,0.11103733380635579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,16384,0.02049688829316033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,12288,0.02306488818592495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,16384,0.019871999820073444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,256,0.0032337777730491427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,65536,0.06763555606206258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,10240,0.02366399930583106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,12288,0.016766221986876596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,65536,0.1397964424557156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,12288,0.03313422203063965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,8192,0.018612444400787354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,10240,0.014800000521871777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,10240,0.02905866503715515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,8192,0.01311644415060679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,7168,0.016954667038387723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,16384,0.041512889994515315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,12288,0.01641688909795549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,7168,0.011720000041855706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,6144,0.01513777838812934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,7168,0.011081777513027191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,10240,0.014581332604090372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,6144,0.011038222246699862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,6144,0.010388444695207808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,5120,0.01332266628742218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,8192,0.012499555945396423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,8192,0.024887111451890733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,4096,0.012084444363911947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,5120,0.009381333159075843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,5120,0.00980444418059455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,5120,0.01886133352915446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,3584,0.01036177741156684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,4096,0.00889511075284746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,51200,0.05325599842601352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,4096,0.016941333810488384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,3584,0.008309333688682979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,3584,0.014480888843536377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,6144,0.021014221840434607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,3072,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,3072,0.007585778004593319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,2560,0.008651555412345463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,3072,0.01347555551264021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,4096,0.00869511150651508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,2048,0.007228444847795699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,16384,0.02688088847531213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,2560,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,2560,0.012774222426944308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,1536,0.005917333480384614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,2048,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,2048,0.006665777828958299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,3072,0.009335111412737105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,1024,0.004602666530344221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,1536,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,1536,0.010416000253624385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,1024,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,768,0.004226666771703296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,1024,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,7168,0.022729777627521094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,768,0.00565422202150027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,768,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,3584,0.008048888709810045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,2048,0.011838222543398539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,1536,0.006055999961164262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,512,0.005384888913896348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,512,0.00612088872326745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,512,0.008664888640244802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,1024,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,256,0.005207111024194294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,2560,0.007666666474607255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,512,0.003581333491537306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,768,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,128,0.007646222081449296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,128,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,64,0.002986666642957263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,256,0.003548444559176763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,64,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,256,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,128,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,65536,0.06853866577148438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,256,0.008033778104517195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,51200,0.05567377805709839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,65536,0.061259554492102734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,65536,0.13992622163560656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,51200,0.05110133356518216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,32,0.0028764444092909494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,16384,0.023725334140989516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,51200,0.11053155528174506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,32,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,12288,0.020191111498408847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,16384,0.04140888982348972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,65536,0.06496000289916992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,12288,0.016342222690582275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,10240,0.017539555827776592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,12288,0.033055110110176936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,51200,0.048617776897218495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,10240,0.01442577772670322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,10240,0.014135999812020196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,10240,0.02904799911710951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,8192,0.012334222594896952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,7168,0.014888889259762235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,128,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,12288,0.016155555844306946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,7168,0.011723555624485016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,7168,0.011051555474599203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,8192,0.01679377754529317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,16384,0.01959733333852556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,6144,0.010477333433098264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,5120,0.011743110915025076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,6144,0.010047111246320937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,6144,0.02102488941616482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,8192,0.024819556209776137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,4096,0.01035377797153261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,5120,0.009398221969604492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,5120,0.018575999471876357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,6144,0.013444444371594323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,4096,0.008207110895050896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,3584,0.010027555955780877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,4096,0.015962666935390897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,16384,0.01995111174053616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,3584,0.008440888590282863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,8192,0.01275111072593265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,3072,0.00738933351304796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,5120,0.00903200027015474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,2560,0.007632888853549957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,3072,0.013600000076823764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,7168,0.022708444131745234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,4096,0.00868000007337994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,2048,0.006612444503439798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,2560,0.007749333149856991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,2560,0.01240355521440506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,3072,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,1536,0.0053280000057485365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,2048,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,2048,0.011577777564525604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,1536,0.0062977779242727495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,1024,0.004583111239804162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,3072,0.008055110772450766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,2560,0.0069520001610120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,1024,0.009385777844323052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,3584,0.00794400026400884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,3584,0.014577777849303352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,768,0.005616888817813661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,1536,0.0069324444565508105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,512,0.003552000141806073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,768,0.00904888908068339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,1024,0.005649777750174205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,512,0.008647999829716152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,256,0.003583111282851961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,2048,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,256,0.0052391112678580815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,1536,0.01071200023094813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,768,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,256,0.00600000015563435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,512,0.005453333258628845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,128,0.004981333182917701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,64,0.0029013332807355454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,768,0.0042275554604000514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,32,0.0032337777730491427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,64,0.005250666704442766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,65536,0.05600000090069241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,256,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,128,0.003185777821474605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,65536,0.062103999985588916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,51200,0.04601777924431694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,128,0.008008889026112026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,1024,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,51200,0.047637333472569786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,51200,0.049238221512900464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,16384,0.020201777418454487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,32,0.0052764444715446895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,16384,0.019362666540675692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,16384,0.041230221589406334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,12288,0.016184000505341422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,512,0.005975111077229182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,12288,0.01555022266176012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,12288,0.015669332610236276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,12288,0.032638221979141235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,10240,0.015602666470739575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,10240,0.014137778017255994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,65536,0.05984089109632704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,10240,0.028554668029149372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,8192,0.014864888456132678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,10240,0.013696889082590738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,128,0.005964444329341252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,7168,0.013740444348918067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,8192,0.011763555308183035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,8192,0.012085333466529846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,6144,0.01276266657643848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,7168,0.010777778095669217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,51200,0.11094578107198079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,7168,0.02238133384121789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,5120,0.010296000374688042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,6144,0.010043555663691627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,6144,0.019938665959570143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,16384,0.019274666905403137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,4096,0.008993777963850234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,5120,0.009054222040706212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,65536,0.1395768854353163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,5120,0.017887999614079792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,3584,0.008355555435021719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,4096,0.00868711123863856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,8192,0.024821332759327356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,7168,0.011336888704035016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,3584,0.007685333490371704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,3584,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,3584,0.014147554834683737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,3072,0.007318221860461765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,2560,0.0069466663731469055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,5120,0.009359999663299983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,4096,0.008010666403505538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,4096,0.015834665960735746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,2048,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,2560,0.007684444387753804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,3072,0.007633777956167857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,6144,0.010072888599501716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,2048,0.007271111011505127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,2048,0.011395555403497485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,3072,0.013373333546850415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,1536,0.006134222365087933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,1536,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,1024,0.004568888909286923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,1536,0.010682666467295753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,2560,0.012102222277058495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,768,0.004135999828577042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,1024,0.00583022211988767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,1024,0.009355555805895064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,768,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,512,0.003552000141806073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,768,0.009005332986513773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,512,0.005294222384691238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,512,0.008707555631796518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,256,0.0032257777121331957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,512,0.005965333431959152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,1024,0.006618666566080517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,256,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,256,0.005091555416584015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,128,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,2048,0.006671999891599019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,3072,0.007976888782448238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,64,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,128,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,32,0.0028657778683635923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,2560,0.006952889263629913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,65536,0.04569688770506117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,32,0.0052826665341854095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,1536,0.0052435555391841466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,65536,0.059559113449520536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,51200,0.036880890528361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,65536,0.13961866166856554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,128,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,65536,0.05829155445098877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,128,0.007646222081449296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,16384,0.016562667157914903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,51200,0.04717600014474657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,51200,0.046759112013710864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,51200,0.11045777797698975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,12288,0.014474666780895658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,16384,0.018965333700180054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,16384,0.018464000688658822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,10240,0.015187554889255099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,12288,0.014464888307783337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,16384,0.04099199838108487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,256,0.007661333514584436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,8192,0.014998222390810648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,10240,0.01348622226052814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,10240,0.013153778182135688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,10240,0.028167999453014795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,7168,0.012574222352769641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,8192,0.01178311142656538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,64,0.0052764444715446895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,8192,0.024365334047211543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,7168,0.01071911139620675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,7168,0.01071999967098236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,7168,0.021507556239763897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,12288,0.015531554818153381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,768,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,5120,0.010025777750545079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,6144,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,6144,0.01958311100800832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,5120,0.009373333719041612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,4096,0.00870400004916721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,6144,0.011133333047231039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,4096,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,4096,0.008663111262851292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,12288,0.03295466634962294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,6144,0.010037333601050908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,4096,0.015497777197096082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,3584,0.008344000412358178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,8192,0.01145600030819575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,3072,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,5120,0.017779555585649278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,3584,0.014168888330459595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,2560,0.006267555471923616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,3072,0.00797333319981893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,3072,0.0073626670572492815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,3072,0.013418667018413544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,2048,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,2560,0.007687110867765214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,2560,0.012132444315486483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,2048,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,1536,0.004871110949251386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,5120,0.00870222184393141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,1536,0.006246222390068903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,1024,0.004193777839342753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,1536,0.010403555300500657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,2560,0.006937777830494775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,1024,0.005624889085690181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,1024,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,768,0.003891555385457145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,2048,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,2048,0.01110400011142095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,768,0.005570666657553778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,768,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,1536,0.0069573331210348345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,512,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,256,0.0032088889016045462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,512,0.006253333141406377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,3584,0.007967111137178209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,1024,0.009369778136412302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,256,0.004969777746333016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,256,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,256,0.008010666403505538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,3584,0.007654222349325816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,768,0.006266666783226862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,128,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,128,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,32,0.0028808888875775864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,128,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,64,0.005037333402368757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,32,0.005250666704442766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,512,0.008354666332403818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,65536,0.05736266904407077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,65536,0.05746399694018894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,128,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,65536,0.13911110825008816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,16384,0.013441777891582914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,51200,0.045315557056003146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,51200,0.046206222640143506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,51200,0.11013688643773396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,12288,0.015154666370815702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,16384,0.01817599932352702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,16384,0.040808889600965716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,12288,0.014408888088332282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,51200,0.02756799923049079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,10240,0.013788444300492605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,10240,0.01311555587583118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,512,0.0035342222286595237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,8192,0.012484444512261285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,16384,0.016914667354689706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,8192,0.01110311100880305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,12288,0.014143110977278816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,8192,0.023728888895776536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,8192,0.01144266708029641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,7168,0.010332444475756751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,64,0.0029066666546795103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,7168,0.021412443783548143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,6144,0.010728889041476779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,7168,0.0105591110057301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,65536,0.03351199958059523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,5120,0.00981955561372969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,6144,0.010055999788973067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,6144,0.009420444567998251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,6144,0.019322666856977675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,4096,0.007970666719807519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,7168,0.011671110987663269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,5120,0.017261332935757108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,4096,0.008352888955010308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,3584,0.011361777782440186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,10240,0.012746666868527731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,3584,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,3584,0.014135999812020196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,12288,0.03186933199564616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,3584,0.008340444829728868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,4096,0.008020444048775567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,10240,0.028016000986099243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,3072,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,2560,0.009001777403884465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,3072,0.013153778182135688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,2560,0.00666133314371109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,3072,0.009356444080670675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,2048,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,4096,0.015207111835479736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,2048,0.0064533332155810455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,2048,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,1536,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,5120,0.00868977771864997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,2560,0.012422222230169507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,2560,0.007646222081449296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,1024,0.0042160000238153665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,1536,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,3072,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,768,0.0038853333228164246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,1024,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,1024,0.005608888963858287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,1024,0.009383999639087254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,512,0.0035377778112888336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,768,0.006243555496136348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,768,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,768,0.009069333473841349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,256,0.003248888792263137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,512,0.005247999810510212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,512,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,256,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,1536,0.010394666757848527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,256,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,128,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,2048,0.011129777464601727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,64,0.0028986665937635633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,128,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,32,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,1536,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,768,65536,0.025798221429189045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,64,0.004936000125275718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,32,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,128,0.0029155556112527847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,768,51200,0.022858666049109563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,65536,0.05594222413169014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,65536,0.13913955953386095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,51200,0.04450666573312548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,51200,0.04593511091338264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,512,0.006050666587220297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,51200,0.1099555558628506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,16384,0.016892444756295946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,768,12288,0.012400888734393649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,256,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,16384,0.04047822289996677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,65536,0.05702222055859036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,12288,0.014135999812020196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,768,10240,0.012595555848545499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,12288,0.03187733226352268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,128,0.007975999679830339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,768,16384,0.014486221803559197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,5120,0.009383111364311641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,10240,0.027450667487250432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,16384,0.017553778158293832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,768,7168,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,8192,0.023623110519515142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,12288,0.014132445057233175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,7168,0.021074665917290583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,7168,0.010738666686746808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,768,6144,0.010079999764760336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,768,8192,0.011443555355072021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,6144,0.01019200020366245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,6144,0.009357333183288574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,8192,0.011390222443474663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,6144,0.01926399932967292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,5120,0.009385777844323052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,5120,0.008683555656009251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,768,4096,0.008016889293988546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,5120,0.017215111189418368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,4096,0.00867377801073922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,4096,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,768,3584,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,4096,0.01554666625128852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,3584,0.00833777752187517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,3584,0.007646222081449296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,8192,0.01108266661564509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,3584,0.014143110977278816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,3072,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,3072,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,3072,0.013408888545301227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,2560,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,2560,0.012153777811262341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,2560,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,768,5120,0.009006222089131674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,768,3072,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,2048,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,768,1536,0.004620444443490771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,768,2560,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,7168,0.010083555347389644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,1536,0.00600355573826366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,1536,0.010436444646782346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,768,1024,0.004228444563017951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,768,2048,0.0052879999081293745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,1536,0.0069466663731469055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,1024,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,1024,0.009695111049546136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,768,768,0.0039048890272776284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,768,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,768,0.009015111459626092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,768,512,0.0035546666218174826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,512,0.0052728888889153796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,2048,0.011424000064531961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,1024,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,768,256,0.0031742221779293488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,512,0.006118222243256039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,10240,0.01254755589697096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,256,0.005028444445795483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,768,0.006314666734801398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,256,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,128,0.005894222193294101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,128,0.00498488876554701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,768,64,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,768,32,0.0028568889117903183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,64,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,512,65536,0.0284897784392039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,10240,0.012771555946932899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,32,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,768,128,0.0028479999552170434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,256,0.005962666538026597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,65536,0.057168887721167676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,65536,0.1388568878173828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,51200,0.04388622111744351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,51200,0.04394488864474826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,128,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,512,0.008567111359702216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,16384,0.016754666964213055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,16384,0.039953778187433876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,512,51200,0.02403555479314592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,65536,0.055013332102033824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,12288,0.01390044391155243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,12288,0.014409777190950183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,512,10240,0.01036977767944336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,12288,0.03153511219554477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,51200,0.10998666948742336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,512,8192,0.009008000294367472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,10240,0.012596444123321109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,10240,0.027433777848879497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,16384,0.01686755485004849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,512,12288,0.011748444702890186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,8192,0.02372088862790002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,2048,0.007354666789372762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,7168,0.021018667353524104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,512,16384,0.016193777322769165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,512,6144,0.008049777812427944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,10240,0.012892444100644855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,6144,0.009357333183288574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,512,5120,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,512,7168,0.008368000388145447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,5120,0.008678221868144141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,7168,0.010040889183680216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,5120,0.01722755531469981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,5120,0.009350222018029954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,7168,0.010793777803579966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,4096,0.008711111214425828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,512,3584,0.007365333537260692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,4096,0.015494222442309061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,6144,0.019204444355434842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,8192,0.01090666651725769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,512,3072,0.006604444649484422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,3584,0.007650666766696506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,3584,0.014150222142537435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,512,4096,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,3072,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,3072,0.013221333424250284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,4096,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,2560,0.006979555719428592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,6144,0.010012444522645738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,2560,0.007646222081449296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,3584,0.008389333056079017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,2048,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,512,1536,0.0047439999050564235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,2048,0.011420444481902652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,3072,0.007297777467303806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,512,2560,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,512,1024,0.003989333493842019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,1536,0.00684533350997501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,1536,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,1536,0.010392888552612728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,512,768,0.0035964445107513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,512,2048,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,2560,0.012111999922328524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,1024,0.00904444439543618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,512,512,0.003528000166018804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,768,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,768,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,2048,0.007344889144102733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,512,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,512,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,512,0.008380444513426887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,256,0.005120888766315249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,1024,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,1024,0.005623111294375525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,256,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,8192,0.011414222419261932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,128,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,128,0.004936000125275718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,128,0.007358222372002072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,512,32,0.0029048888633648553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,512,256,0.0029013332807355454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,256,65536,0.018606222338146634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,64,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,32,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,256,0.00798044436507755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,512,128,0.0029013332807355454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,256,51200,0.01623644431432088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,65536,0.13852977752685547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,768,0.006353777729802662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,512,64,0.0028417776856157514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,51200,0.1096915536456638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,16384,0.01683644453684489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,256,12288,0.008678221868144141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,16384,0.039515554904937744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,65536,0.053951111104753285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,65536,0.04969244533114963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,12288,0.013778666655222574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,12288,0.014263111684057446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,12288,0.031565331750445895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,51200,0.04018044471740723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,256,8192,0.01074577785200543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,10240,0.012421333127551608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,10240,0.012847999731699625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,10240,0.02737066646416982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,256,7168,0.010064000056849586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,16384,0.016862221890025668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,8192,0.01073333372672399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,8192,0.023696889479955036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,256,6144,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,7168,0.010737777584128909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,7168,0.021136000752449036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,256,10240,0.012763555679056378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,256,5120,0.007977777885066139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,6144,0.009291555318567488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,256,16384,0.010057777994208867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,6144,0.018943111101786297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,256,4096,0.007663999994595845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,5120,0.008455110920800103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,5120,0.01753155555990007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,8192,0.011422221859296164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,256,3584,0.007332444190979004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,4096,0.008634666601816813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,4096,0.015497777197096082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,7168,0.010106667048401302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,256,3072,0.006994666324721442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,3584,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,3584,0.014190221826235453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,6144,0.010046222143703038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,51200,0.04286488890647888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,3072,0.007295110987292395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,3072,0.007984889050324758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,3072,0.013328889177905189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,256,2048,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,2560,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,4096,0.007828444242477417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,2048,0.006393777827421824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,2048,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,256,1536,0.004567111117972268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,2048,0.011401777466138205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,3584,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,256,1024,0.003887999802827835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,1536,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,1536,0.005971555494599872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,256,2560,0.0058986664646201665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,1536,0.01036711119943195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,1024,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,1024,0.005595555735958948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,2560,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,1024,0.009352888498041365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,768,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,2560,0.0124177775449223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,768,0.009000000026490953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,256,256,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,512,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,512,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,256,0.0058959999846087555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,256,128,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,5120,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,256,512,0.003185777821474605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,128,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,256,64,0.002536000063021978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,128,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,128,0.008014221986134848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,512,0.006301333506902059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,128,65536,0.01514933341079288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,64,0.004943111290534337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,32,0.005045333256324132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,256,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,256,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,128,51200,0.013170666992664337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,65536,0.0493253337012397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,65536,0.13772177696228027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,768,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,51200,0.03956266575389438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,51200,0.10824355814192031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,256,32,0.002879111096262932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,16384,0.016865778300497267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,128,12288,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,16384,0.039464887645509504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,256,768,0.0037271111375755737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,65536,0.05205066667662727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,128,10240,0.00907022257645925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,12288,0.013791110780504016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,12288,0.014148443937301636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,12288,0.03115733464558919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,51200,0.0414373344845242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,10240,0.012774222426944308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,10240,0.012435555458068848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,10240,0.027089777919981215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,128,7168,0.007915555602974361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,8192,0.01145333300034205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,8192,0.010601777997281818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,16384,0.016582222448454965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,8192,0.023375110493765935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,7168,0.010704889065689512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,7168,0.010045333041085137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,7168,0.02103288968404134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,128,5120,0.00701333334048589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,6144,0.01924533314175076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,128,8192,0.008382221890820397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,128,16384,0.009681777821646797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,128,4096,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,5120,0.00868888861603207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,5120,0.017282666431532968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,4096,0.0079644446571668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,128,3584,0.006505777852402792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,6144,0.010009778042634329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,3584,0.007540444533030192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,3584,0.013984888792037964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,128,3072,0.005998222364319696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,5120,0.009015111459626092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,3072,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,128,6144,0.007305777735180325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,4096,0.008575999902354347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,6144,0.009301332963837517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,2560,0.01218488895230823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,2560,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,2048,0.0063235556913746735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,3072,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,128,2560,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,128,1536,0.0042275554604000514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,2048,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,2048,0.011420444481902652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,2560,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,128,1024,0.0038746665749284956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,3584,0.00831822223133511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,128,2048,0.005985777825117111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,1024,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,128,768,0.0035377778112888336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,1024,0.009691555466916826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,4096,0.01555377741654714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,768,0.005608888963858287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,128,512,0.0032586666444937387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,768,0.009337777892748514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,1536,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,1536,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,128,256,0.0028737777223189673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,512,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,512,0.00868266655339135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,1024,0.005569777968857024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,128,128,0.002571555475393931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,256,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,3072,0.013447999954223633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,256,0.007620444728268518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,128,64,0.002611555573013094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,128,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,128,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,128,32,0.002568888788421949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,128,0.007646222081449296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,768,0.006343110981914733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,64,0.004902222090297275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,32,0.00462755560874939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,64,16384,0.009026666482289633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,65536,0.05083822210629781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,64,12288,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,51200,0.040778666734695435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,64,10240,0.0074728892909155945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,256,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,64,8192,0.006984000404675801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,12288,0.01367644468943278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,64,7168,0.006948444578382704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,10240,0.012113778127564324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,64,6144,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,8192,0.01071111112833023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,64,5120,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,7168,0.010020444790522257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,1536,0.01038755559258991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,64,4096,0.005969777703285217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,6144,0.009012444151772393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,5120,0.008553778131802877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,64,3072,0.00599911105301645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,16384,0.016527111331621807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,3584,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,64,2560,0.006592888798978593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,3072,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,64,2048,0.005962666538026597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,2560,0.0063991112013657885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,64,51200,0.01402844488620758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,64,1536,0.004558222161398994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,64,1024,0.003721777763631609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,64,3584,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,2048,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,1536,0.005984888722499211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,64,512,0.0032435556252797446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,4096,0.007670222057236566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,64,256,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,768,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,64,128,0.0027991111079851785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,512,0.005327111317051782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,64,64,0.002536000063021978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,256,0.004965333475006951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,64,32,0.002565333412753211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,128,0.00489777781897121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,32,65536,0.01482399966981676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,512,0.005236444373925527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,64,0.004956444518433677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,32,51200,0.012072888513406118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,64,768,0.003532444437344869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,1024,0.005590222362014983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,32,12288,0.00757422215408749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,16384,0.016559999850061204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,32,10240,0.007626666790909237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,64,65536,0.018030222919252183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,12288,0.013760888742076026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,32,8192,0.0070293330483966405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,32,7168,0.006695111178689533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,32,6144,0.006343110981914733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,8192,0.010545777777830759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,32,16384,0.008683555656009251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,32,5120,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,32,4096,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,65536,0.05079999897215101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,51200,0.04110488957828946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,5120,0.008384000096056197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,32,3072,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,3584,0.0074346669846110875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,32,2560,0.0063866666621632045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,10240,0.01238666640387641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,3072,0.0070159998204973005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,32,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,7168,0.010048888623714447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,32,1536,0.004385777645640903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,32,1024,0.0038951109680864546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,2048,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,32,768,0.0033004445334275565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,1536,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,4096,0.007648000286685095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,768,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,32,256,0.0028417776856157514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,512,0.004987555659479565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,32,2048,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,256,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,32,64,0.00249422217408816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,128,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,32,32,0.002520888836847411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,64,0.005032888717121548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,1024,0.005631999837027655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,32,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,32,128,0.0028702221396896574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,2560,0.00664533343580034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,6144,0.009408888717492422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,16384,0.3379937807718913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,51200,1.0435849295722115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,51200,0.8318693372938367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,32,512,0.0032195556494924757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,12288,0.25768976741366917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,16384,0.1362933317820231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,12288,0.10199466678831313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,12288,0.20213688744439018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,8192,0.17599822415245903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,10240,0.08667111396789551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,160,32,3584,0.006180444525347815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,7168,0.16187111536661783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,8192,0.13792267110612658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,16384,0.26538311110602486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,6144,0.13312444421980116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,7168,0.12070400185055202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,10240,0.21726489067077637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,5120,0.1150124470392863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,6144,0.1016133361392551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,10240,0.1669315497080485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,5120,0.04583555459976196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,4096,0.03925688730345832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,8192,0.07003733184602526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,3584,0.08337244722578262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,7168,0.06264355447557238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,3584,0.034894221358829074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,3584,0.06339555316501193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,6144,0.05375733309321933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,2560,0.0609075559510125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,5120,0.0858115553855896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,3072,0.030745777818891738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,3072,0.054634667105144925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,2048,0.050629334317313306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,4096,0.07036622365315755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,2048,0.023576888773176406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,2048,0.040024889840020075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,1536,0.03974399964014689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,3072,0.0733955568737454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,51200,0.41351821687486434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,1024,0.029939555459552344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,1536,0.018698667486508686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,1536,0.031707555055618286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,4096,0.09619910849465264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,1024,0.024506666594081458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,512,0.01585244470172458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,2560,0.027261333333121404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,512,0.013065777719020844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,256,0.01091733326514562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,768,0.022952889402707417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,1024,0.015547555353906421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,256,0.01571555601225959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,768,0.014184888866212634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,2560,0.04712444543838501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,128,0.01478488908873664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,32,0.008344000412358178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,512,0.0222746663623386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,64,0.014576888746685453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,32,0.015265777707099915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,256,0.011700444751315646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,128,0.011735999749766456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,768,0.03087644444571601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,65536,0.8448853492736816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,51200,0.890718248155382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,16384,0.28822755813598633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,51200,0.643937799665663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,65536,0.4113173219892714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,64,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,12288,0.21822844611273873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,65536,1.1313031514485676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,12288,0.16740711530049643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,12288,0.08083466688791911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,10240,0.1986080010732015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,8192,0.1608088943693373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,128,0.00904266701804267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,10240,0.0687991115781996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,10240,0.1427839994430542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,8192,0.11119200123680963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,7168,0.09631022479799058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,16384,0.1070222192340427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,7168,0.050813333855734936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,16384,0.21660889519585502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,5120,0.10559378067652385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,6144,0.08247111241022746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,5120,0.037676443656285606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,5120,0.06953955358929105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,4096,0.08409777614805435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,3584,0.07422044542100695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,7168,0.14212889141506618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,8192,0.055832001898023814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,4096,0.0595760014322069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,3072,0.06473422050476074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,6144,0.12265066305796306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,3584,0.02944177720281813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,2560,0.054457777076297335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,3584,0.050901333491007485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,2048,0.04457866814401415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,3072,0.02606044378545549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,6144,0.04369866516855028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,1536,0.03458755546145969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,2048,0.019888889458444383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,4096,0.03276711040072971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,1024,0.025892444782786902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,1536,0.016303999556435477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,1536,0.026520889666345384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,768,0.019280888968043856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,1024,0.020210666788948905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,3072,0.044864889648225575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,768,0.0127697777416971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,768,0.024599111742443506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,512,0.011633777783976661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,256,0.009859555297427708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,256,0.013249778085284762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,256,0.01054488867521286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,128,0.006601777755551868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,2560,0.023306666149033442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,512,0.013968888256284924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,64,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,128,0.010609777437316047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,51200,0.3177564409044054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,128,0.012437333663304647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,64,0.012444444828563266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,65536,0.3404160075717502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,65536,0.2977004316118029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,65536,0.13793067137400308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,32,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,2048,0.032955557107925415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,16384,0.09027644660737778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,32,0.01309866706530253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,51200,0.11047289106580947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,12288,0.07011822197172377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,16384,0.040024889840020075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,2560,0.038378665844599404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,51200,0.26660090022616917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,12288,0.03190577692455716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,10240,0.057187557220458984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,51200,0.23330489794413248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,10240,0.050829334391487964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,10240,0.027607111467255488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,16384,0.0779013302591112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,7168,0.0421324438518948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,8192,0.0428017775217692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,512,0.018239110708236694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,6144,0.03712800145149231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,1024,0.013855111267831592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,8192,0.04907110995716519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,6144,0.03312355611059401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,6144,0.019016888406541612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,8192,0.023423999547958374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,4096,0.0266977780395084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,5120,0.028592887851927016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,12288,0.059327112303839795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,3584,0.023642667465739783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,3584,0.021553777986102637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,5120,0.031630221340391375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,3584,0.013739555246300168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,3072,0.013047110703256396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,2560,0.017553778158293832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,3072,0.019208000765906442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,7168,0.03747911254564921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,2560,0.012070222033394707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,2048,0.014497778481907315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,2560,0.016926222377353244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,1536,0.011751111182901593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,2048,0.010999110837777456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,2048,0.01482222146458096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,1024,0.009287111461162567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,1536,0.009166222479608323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,1536,0.012717333104875354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,768,0.00797333319981893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,5120,0.016905777984195285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,1024,0.01036888857682546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,512,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,768,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,768,0.011415111521879831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,256,0.004568888909286923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,4096,0.01514844430817498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,512,0.0073626670572492815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,128,0.0038888889054457345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,256,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,256,0.007833778030342525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,64,0.0035253332720862497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,3072,0.020925333102544148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,128,0.007345777418878343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,32,0.004234666625658671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,1024,0.008013333711359236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,64,0.007286222444640265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,65536,0.27500976456536186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,32,0.00793333351612091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,7168,0.02131555477778117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,4096,0.024156444602542456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,512,0.009411555197503831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,65536,0.1258266634411282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,51200,0.16367822223239475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,128,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,16384,0.03722577624850803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,16384,0.05702577696906196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,10240,0.047219554583231606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,12288,0.02960977620548672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,12288,0.04367466767628988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,8192,0.04057777921358744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,10240,0.025399110383457605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,10240,0.03755199909210205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,51200,0.2145235538482666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,7168,0.03497955534193251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,8192,0.021818665994538203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,16384,0.07449155383639865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,6144,0.03030666708946228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,7168,0.019880889190567862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,5120,0.026150221625963848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,12288,0.05642311440573799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,6144,0.024953777591387432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,4096,0.022687110635969374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,5120,0.016206221448050607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,5120,0.021334222621387903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,3584,0.01959288948112064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,4096,0.014641778336630927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,4096,0.018114666144053142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,3072,0.017380444539917838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,3584,0.013562666873137156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,3584,0.01648533344268799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,8192,0.032171554035610624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,2560,0.014793778459231058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,51200,0.10007911258273655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,3072,0.015238222148683337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,2048,0.01311911145846049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,7168,0.028148445818159316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,1536,0.010685332947307162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,2560,0.013463999662134381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,2048,0.010002666877375709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,1024,0.00868088917599784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,2048,0.012088889049159156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,1536,0.008654221892356873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,1536,0.01071111112833023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,768,0.007627555893527136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,1024,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,1024,0.00869066682126787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,512,0.005912888795137405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,768,0.007676444119877285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,3072,0.01253600004646513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,768,0.009705777797434065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,256,0.004563555535342959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,512,0.007263110743628607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,6144,0.018057778477668762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,128,0.003923555629120933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,256,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,256,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,128,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,32,0.003896000070704354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,64,0.006657777975002925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,32,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,65536,0.21142844359079996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,65536,0.2661493354373508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,2560,0.011615999870830111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,512,0.007939555578761632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,65536,0.20047911008199057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,65536,0.12338577376471625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,64,0.003624000069167879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,16384,0.0699288911289639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,128,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,12288,0.05496889021661547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,51200,0.0980444418059455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,51200,0.15688888231913248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,10240,0.046111110183927745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,16384,0.05433600147565206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,12288,0.04191822144720289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,8192,0.03896266553137038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,10240,0.024694222542974684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,10240,0.0359280010064443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,7168,0.034041778908835515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,6144,0.029822223716311987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,8192,0.03050311075316535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,51200,0.210190216700236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,5120,0.025735111700163946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,6144,0.017629333668284945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,6144,0.023623110519515142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,4096,0.021568889419237774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,16384,0.035848889085981585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,5120,0.02036355601416694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,12288,0.028479999966091577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,3584,0.018981332580248516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,4096,0.017595556047227647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,3072,0.01685333251953125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,3584,0.013150222599506378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,8192,0.02142133315404256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,7168,0.019319999549123976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,3072,0.014728888869285583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,2560,0.010701333483060202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,2560,0.013117333253224691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,2048,0.012481778032249875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,5120,0.015841777125994366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,1536,0.010482666393121084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,2048,0.009003555609120263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,4096,0.014206222361988492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,2048,0.011699555648697747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,1024,0.008440000315507254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,1536,0.008419555922349295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,1536,0.010253333383136326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,3584,0.015953777564896476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,2560,0.014499555031458536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,512,0.005792888916201062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,768,0.007333333293596904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,768,0.009047110875447592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,256,0.004564444637960858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,512,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,512,0.007866666548781926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,256,0.006980444822046492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,128,0.003899555653333664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,256,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,64,0.0035413333939181436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,128,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,32,0.0035786665976047516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,768,0.006953777538405524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,1024,0.00775911079512702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,3072,0.01202311118443807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,64,0.006579555571079254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,65536,0.19621600045098198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,7168,0.02697155541843838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,32,0.0070088886552386815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,51200,0.1539191140068902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,16384,0.053073777092827686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,65536,0.07401511404249403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,51200,0.0599235561158922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,12288,0.04082399937841628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,16384,0.02346666653951009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,16384,0.043354666895336576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,12288,0.01920711166328854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,12288,0.03390222125583225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,128,0.006304889089531369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,8192,0.029530667596393164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,10240,0.029139555162853662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,8192,0.01519733336236742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,8192,0.02472355630662706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,65536,0.1576204432381524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,51200,0.1227102279663086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,7168,0.02163999941613939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,7168,0.013916444447305469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,10240,0.03447555502255758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,5120,0.019302222463819716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,6144,0.012852444416946836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,6144,0.01926399932967292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,5120,0.011861333416567909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,5120,0.016869333055284288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,7168,0.025639999243948195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,3584,0.014468444718254937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,4096,0.014904000692897372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,3584,0.009735999835862054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,6144,0.022466666168636743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,3072,0.01313244468635983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,10240,0.01721600029203627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,3072,0.00871288859181934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,2560,0.011395555403497485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,2560,0.008062221937709385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,2048,0.010068444742096795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,2048,0.010391111175219217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,1536,0.008419555922349295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,3072,0.012514666550689273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,4096,0.016181333197487723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,1536,0.008575999902354347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,1536,0.007329777710967594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,4096,0.010982222027248807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,1024,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,768,0.005444444302055571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,768,0.00665244460105896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,3584,0.01370666672786077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,512,0.004573333180612988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,512,0.0063440000845326324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,256,0.003908444609906939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,1024,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,1024,0.006887110984987683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,128,0.003559111307064692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,256,0.006079111248254776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,2560,0.011639110743999481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,1024,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,128,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,32,0.003555555724435382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,2048,0.007688889073001013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,768,0.008168000313970778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,64,0.005965333431959152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,65536,0.16562044620513916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,32,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,512,0.0070222218831380205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,65536,0.0714062253634135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,65536,0.1534471114476522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,16384,0.05015110969543457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,51200,0.05794488721423679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,256,0.006186666587988536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,64,0.003228444399105178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,16384,0.042565332518683545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,10240,0.0332951115237342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,12288,0.018723555737071566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,12288,0.0328764451874627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,8192,0.028243554963005915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,10240,0.016602666841612924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,10240,0.02795555525355869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,7168,0.024696888195143804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,51200,0.13218400213453504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,8192,0.023777777949968975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,6144,0.021700445148679946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,7168,0.013626666532622443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,7168,0.020946666598320007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,128,0.006250666661394968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,5120,0.018570666511853535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,6144,0.012484444512261285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,4096,0.016191111670600045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,5120,0.01676177812947167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,4096,0.010047111246320937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,3584,0.014161777165200977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,51200,0.12008178234100342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,3584,0.013449778159459433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,3072,0.01276533305644989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,3584,0.009076444639099969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,6144,0.018927110566033255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,2560,0.011074666347768573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,3072,0.012427555190192329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,4096,0.014761777387724983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,2560,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,2048,0.010029333333174387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,2048,0.007607111500369177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,1536,0.008372444245550368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,12288,0.03977333174811469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,1536,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,5120,0.01143022212717268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,1024,0.006606222026877933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,3072,0.00849955528974533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,768,0.0052435555391841466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,1024,0.006654222392373615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,1024,0.007291555404663086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,8192,0.014536889062987434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,768,0.008045333127180735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,2048,0.009856888817416297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,512,0.006796444455782573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,256,0.003866666720973121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,1536,0.008344000412358178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,256,0.005990222096443176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,128,0.0035573335157500375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,512,0.004580444345871608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,128,0.005995555470387141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,64,0.0032844444115956626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,16384,0.022964444425370958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,32,0.003564444267087512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,64,0.006072000082996156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,256,0.005991999887757831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,2560,0.011433777709801992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,65536,0.11549599965413411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,512,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,65536,0.06901244322458903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,51200,0.12112889024946426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,16384,0.046560890144772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,51200,0.05615466833114624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,51200,0.08948355250888401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,12288,0.03595644566747878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,32,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,16384,0.021822222405009802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,10240,0.03007022208637661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,12288,0.01797422269980113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,768,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,8192,0.025721778472264607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,10240,0.01605866683853997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,7168,0.022298667165968154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,10240,0.02166755497455597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,8192,0.01885777711868286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,7168,0.01328888866636488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,6144,0.018711111611790128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,7168,0.01683022247420417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,5120,0.016855110724767048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,6144,0.012115555504957834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,6144,0.01534933348496755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,4096,0.014257777068350049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,5120,0.013743999931547375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,12288,0.025388444463411968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,3584,0.012735111018021902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,4096,0.009713778065310584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,128,0.005983110931184556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,4096,0.0123857781291008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,3072,0.011720889144473605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,3584,0.00867022242810991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,2560,0.010695999695195092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,3584,0.01145155562294854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,3072,0.008403555386596257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,2048,0.009687110781669617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,3072,0.01070666644308302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,2560,0.007661333514584436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,1536,0.007916444705592262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,65536,0.15292533238728842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,5120,0.010785777535703448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,1024,0.006268444574541516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,1536,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,1536,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,768,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,1024,0.00666844430896971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,8192,0.014140443669425117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,1024,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,768,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,768,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,256,0.0038862224254343244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,512,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,2560,0.009517333573765224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,2048,0.0075786668393347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,16384,0.03275377882851495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,128,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,64,0.003206222214632564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,512,0.004563555535342959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,32,0.003195555466744635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,64,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,512,0.006304889089531369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,65536,0.12059466044108073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,32,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,128,0.003216000066863166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,2048,0.008271111382378472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,65536,0.11278400156233047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,16384,0.038146668010287814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,128,0.005647111270162795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,16384,0.020987555384635925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,16384,0.03154044349988302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,12288,0.03372800019052293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,12288,0.01759288873937395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,12288,0.02442933287885454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,10240,0.028220444917678833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,51200,0.09630844328138564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,65536,0.0663999981350369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,8192,0.024197333388858374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,10240,0.015881778465376962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,10240,0.021282666259341772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,51200,0.05377599928114149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,8192,0.018221333622932434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,7168,0.012842666771676807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,7168,0.016533333394262526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,6144,0.011614222493436603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,5120,0.01593955523437924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,256,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,256,0.005635555419656966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,5120,0.01349688900841607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,7168,0.021149333980348375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,51200,0.08921688795089722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,4096,0.009053332938088311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,3584,0.012170666621790992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,6144,0.015152000718646579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,3584,0.011035555766688453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,3072,0.011353777514563667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,3584,0.008361777497662438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,5120,0.010416000253624385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,3072,0.008044444852405125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,2560,0.009561777942710454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,3072,0.010033778018421596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,2048,0.008817777865462834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,2560,0.007895111209816402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,2560,0.009018667042255402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,2048,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,2048,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,1536,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,1024,0.00591111100382275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,4096,0.013373333546850415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,8192,0.013794666363133324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,1024,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,1024,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,6144,0.018002667360835604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,512,0.004237333519591225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,4096,0.012109333442317115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,768,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,256,0.003544888976547453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,512,0.006329777754015393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,256,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,256,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,1536,0.007335999773608313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,64,0.0031964443624019623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,128,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,128,0.005631111148330901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,768,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,64,0.005616888817813661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,65536,0.10400088628133138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,1536,0.007158222297827403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,65536,0.0643706652853224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,128,0.003560888684458203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,51200,0.051679111189312406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,32,0.003967111309369405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,16384,0.034655110703574285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,16384,0.020440000626775954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,32,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,768,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,51200,0.08452888992097642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,512,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,12288,0.02349777850839827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,10240,0.015276443627145557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,8192,0.018664000762833487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,16384,0.029726223813162908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,8192,0.013460444079505073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,65536,0.1058373318778144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,12288,0.0169831116994222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,10240,0.02202044427394867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,6144,0.014180445008807711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,7168,0.012080888781282636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,7168,0.016190222567982145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,10240,0.020627554919984605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,5120,0.012480888929631976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,6144,0.01094133324093289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,6144,0.014650666051440768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,5120,0.009387555221716562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,8192,0.0177208897140291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,3584,0.010144000252087912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,4096,0.011744000017642975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,7168,0.01638488968213399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,3072,0.008660444782839881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,3584,0.01073155552148819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,3072,0.00797333319981893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,3072,0.009471110999584198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,4096,0.011403555671374003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,12288,0.027837332752015855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,2560,0.00888888869020674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,2048,0.00664444433318244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,2048,0.007336888876226213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,3584,0.008375110725561777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,51200,0.08327022525999281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,2560,0.008017777568764156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,1024,0.004568888909286923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,1536,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,5120,0.013150222599506378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,768,0.004520888957712385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,1024,0.0063866666621632045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,768,0.006317333214812809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,1536,0.005618666609128316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,768,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,512,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,256,0.003572444534964032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,2560,0.007648000286685095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,256,0.00562755556570159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,4096,0.008683555656009251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,128,0.003216888962520493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,256,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,64,0.0031928889867332247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,128,0.005919999960396025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,2048,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,128,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,64,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,1536,0.007358222372002072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,65536,0.08783466948403253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,1024,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,65536,0.10375910997390747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,512,0.00388711111413108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,32,0.0031848889258172777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,51200,0.08141777912775676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,51200,0.050663110282686025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,512,0.00600888869828648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,32,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,12288,0.02695200012789832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,16384,0.02937777837117513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,10240,0.023734221855799358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,12288,0.016885333591037326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,51200,0.0715893308321635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,8192,0.01721244388156467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,10240,0.02017155620786879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,16384,0.03391733434465196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,7168,0.015646222564909194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,16384,0.0199635558658176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,7168,0.015847111741701763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,65536,0.06321422259012859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,12288,0.023181334137916565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,5120,0.012237333589129977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,6144,0.014482667048772177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,10240,0.01517955627706316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,4096,0.011032889286677042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,5120,0.009518221848540837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,8192,0.013160000244776407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,8192,0.017216889394654166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,4096,0.008653333617581261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,4096,0.011149333582984077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,3584,0.00832622249921163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,3072,0.009128000173303816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,6144,0.013893332746293811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,3584,0.010048888623714447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,2560,0.007630222373538547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,3072,0.008026666939258575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,2048,0.006528000036875407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,2560,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,2560,0.008325333396593729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,1536,0.00554044461912579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,5120,0.013004444539546967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,3584,0.009739555418491364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,1024,0.004564444637960858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,1536,0.006959111326270633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,1024,0.006607111129495833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,768,0.0041271112859249115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,6144,0.01037688884470198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,768,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,768,0.0063075555695427795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,2048,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,2048,0.00792622235086229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,256,0.003548444559176763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,512,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,256,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,256,0.005643555687533484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,128,0.0031973332580592898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,128,0.005998222364319696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,64,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,64,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,32,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,1024,0.0063635553750726916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,7168,0.011745777395036487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,32,0.005399111244413588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,65536,0.07584710915883382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,51200,0.06206488609313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,128,0.0052968888647026485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,65536,0.06547022528118558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,16384,0.026239110363854304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,51200,0.04945066571235657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,51200,0.05122311247719658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,12288,0.022035555707083807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,16384,0.019893333315849304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,3072,0.009365333451165093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,12288,0.016523554921150208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,10240,0.01906399925549825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,10240,0.0148044443792767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,8192,0.015359110302395292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,10240,0.014335110783576965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,8192,0.012431110772821637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,7168,0.014051554931534661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,8192,0.012768000364303589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,7168,0.01107022249036365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,6144,0.012293332980738746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,512,0.0038524443904558816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,6144,0.010097777677906884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,5120,0.011049778097205691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,5120,0.009720889230569204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,5120,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,4096,0.009673777553770278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,12288,0.0162435554795795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,4096,0.00868888861603207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,3584,0.00902133352226681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,3584,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,3584,0.00833155545923445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,3072,0.00830577810605367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,16384,0.019711999429596793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,4096,0.008527999950779809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,3072,0.007648000286685095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,2560,0.007676444119877285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,3072,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,2560,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,2048,0.006711111300521427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,2560,0.00700444479783376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,65536,0.061793777677747935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,7168,0.011744889120260874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,1536,0.005614222337802251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,1536,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,6144,0.010849778023031024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,1024,0.004570666700601578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,1536,0.006214222146405115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,768,0.0042133331298828125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,1024,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,1024,0.005639111002286275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,512,0.003879111260175705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,768,0.005856888989607493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,512,0.006008000009589725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,512,0.006274666637182236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,256,0.003546666767862108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,128,0.0029208889851967492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,256,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,256,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,64,0.003198222153716617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,128,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,32,0.0029786665820413162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,65536,0.06656622224383883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,64,0.005247999810510212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,32,0.00535022219022115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,51200,0.05422133207321167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,65536,0.060714668697781034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,2048,0.006583999842405319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,16384,0.024711110525661047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,51200,0.048206223381890186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,51200,0.049345778094397656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,128,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,12288,0.020991999242040846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,16384,0.01920622256067064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,10240,0.01805688937505086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,768,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,12288,0.015558222929636637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,8192,0.014485332700941296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,10240,0.014091556270917257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,10240,0.013954665925767688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,7168,0.013031110995345645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,8192,0.012192000117566852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,7168,0.011095999843544431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,7168,0.011422221859296164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,16384,0.0195448895295461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,12288,0.01624000072479248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,5120,0.010384000009960597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,6144,0.01038755559258991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,5120,0.009367999931176504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,2048,0.007333333293596904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,4096,0.009167110754383935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,4096,0.008681777450773451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,3584,0.008606221940782335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,4096,0.008394666843944127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,6144,0.011725333001878528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,3072,0.00793422261873881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,3584,0.008375999828179678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,6144,0.010263111028406356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,2560,0.007056000332037608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,3072,0.008042666647169325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,3072,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,2048,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,2560,0.007727999654081132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,512,0.005614222337802251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,5120,0.009375111096435124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,1536,0.0052435555391841466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,2048,0.007344000041484833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,2048,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,1536,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,1536,0.006262222097979651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,768,0.003896000070704354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,1024,0.005629333357016246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,3584,0.007978666987684038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,512,0.0036186666952239144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,768,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,768,0.005720888988839255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,256,0.0032586666444937387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,2560,0.006979555719428592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,1024,0.004274666723277834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,128,0.0031999999450312722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,1024,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,128,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,128,0.005172444300519096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,64,0.002921777880854077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,32,0.0029120000286234748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,512,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,64,0.0052879999081293745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,32,0.005233777893914117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,256,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,256,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,65536,0.05905422237184313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,65536,0.06064266628689236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,51200,0.0470728874206543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,51200,0.047353777620527476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,16384,0.018851555056042142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,65536,0.056516442033979625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,12288,0.016618667377365958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,16384,0.019193778435389202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,51200,0.04653333293067085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,10240,0.01479644411140018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,12288,0.015180443723996481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,12288,0.015423110789722867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,8192,0.014463999205165438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,10240,0.013737777868906656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,10240,0.013447110851605734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,7168,0.01351466692156262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,8192,0.011980444192886353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,8192,0.011748444702890186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,6144,0.011431111229790581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,7168,0.010727110836240979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,7168,0.010833777487277985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,5120,0.010067555639478896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,6144,0.010032888915803697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,6144,0.01000355515215132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,4096,0.009019555317031013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,5120,0.009365333451165093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,5120,0.00902844468752543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,16384,0.020125332805845473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,3584,0.00814844419558843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,4096,0.008658666577604082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,3072,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,4096,0.008021333151393467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,3584,0.007689778175618913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,3072,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,3072,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,2048,0.005921777751710679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,2560,0.007660444411966536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,2560,0.006959111326270633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,1536,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,2048,0.007101332975758447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,1536,0.006263999889294307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,8192,0.011693333586057028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,65536,0.06300444735421075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,2560,0.006488889041874144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,768,0.0038924444880750445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,1024,0.006664888726340399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,768,0.006255110932721033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,768,0.005623999983072281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,2048,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,512,0.005375999957323074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,3584,0.008336000144481659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,512,0.005487111293607288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,1024,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,128,0.0029404444826973807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,512,0.00378933341966735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,128,0.004990222139490975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,64,0.0028453332682450614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,1536,0.007053333024183909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,1024,0.004277333203289243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,64,0.005285333428117964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,32,0.005117333183685939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,65536,0.057631108495924205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,128,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,51200,0.03593511051601834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,512,0.006239111224810283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,256,0.00318755561278926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,16384,0.017190222938855488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,51200,0.046577778127458364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,51200,0.0455040004518297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,65536,0.04447288976775276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,16384,0.018227555685573153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,12288,0.014112000664075216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,12288,0.015029332704014249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,65536,0.057842665248446994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,32,0.0029031110720502008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,10240,0.013658666776286231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,10240,0.012986666626400419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,8192,0.011417778001891242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,8192,0.01144444445768992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,10240,0.013160000244776407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,7168,0.010424888796276517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,6144,0.013053333593739403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,7168,0.01071999967098236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,8192,0.012129777835475074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,5120,0.01015822258260515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,6144,0.009586667021115622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,256,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,4096,0.008996444443861643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,5120,0.009473777479595607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,5120,0.009135999613338048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,4096,0.007991111112965478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,3584,0.007755555212497711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,3584,0.008117333054542542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,3072,0.006677333265542984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,256,0.005103999955786599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,7168,0.014303111367755465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,2560,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,3072,0.007695111135641734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,2560,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,4096,0.008368000388145447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,6144,0.010045333041085137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,2048,0.006610666712125142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,2048,0.007134222322040134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,12288,0.014856000741322836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,3072,0.007299555672539606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,1536,0.006027555714050929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,1024,0.004239999999602635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,2048,0.005711110929648082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,1024,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,768,0.0038693332009845306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,16384,0.018241778016090393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,768,0.005619555711746216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,768,0.006312888943486744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,1536,0.007070221834712558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,2560,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,256,0.0032177778581778207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,512,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,1024,0.005632888939645555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,256,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,128,0.0028693332440323303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,256,0.005030222237110138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,64,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,1536,0.004896000027656555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,32,0.0029057777590221832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,128,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,65536,0.032842665910720825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,512,0.003544888976547453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,32,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,51200,0.027191110783153113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,65536,0.055268443293041654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,16384,0.014465777410401238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,51200,0.046105778879589505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,3584,0.007674666742483775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,12288,0.012878222597969903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,16384,0.016876444220542908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,128,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,10240,0.012089778151777057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,12288,0.014283556077215405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,12288,0.014399111270904541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,64,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,10240,0.012864888542228274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,10240,0.012427555190192329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,65536,0.056962668895721436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,7168,0.010384000009960597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,512,0.006037333359320958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,6144,0.009661333428488838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,51200,0.043488889932632446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,7168,0.010029333333174387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,5120,0.009666666388511658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,6144,0.00998755544424057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,16384,0.01791733337773217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,6144,0.009431111315886179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,4096,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,5120,0.00906577789121204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,3584,0.007536888950400882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,8192,0.010847110715177325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,4096,0.00867022242810991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,3072,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,4096,0.008012444608741337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,3584,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,8192,0.011476444701353708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,2560,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,8192,0.010741333166758219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,3072,0.008103999826643202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,7168,0.010616000327799054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,2560,0.00739644467830658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,1536,0.004905777672926585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,2048,0.007050666544172499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,2560,0.007096000015735626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,1536,0.006618666566080517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,1024,0.004283555679851108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,2048,0.006361777583758037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,5120,0.008664888640244802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,768,0.004023999803596073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,1024,0.006717333363162146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,1024,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,512,0.003626666549179289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,768,0.0063440000845326324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,256,0.0033173333439562055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,2048,0.005487111293607288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,512,0.006081777728266186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,128,0.003031111011902491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,256,0.005889777921968036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,128,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,256,0.005072888814740711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,64,0.0025617778301239014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,32,0.002537777854336633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,128,0.005140444470776452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,768,65536,0.029401779174804688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,3584,0.007639110916190677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,32,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,64,0.0053119998839166425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,3072,0.007280888656775157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,512,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,768,16384,0.01260088880856832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,51200,0.045920888582865395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,65536,0.05395199855168661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,16384,0.016878222425778706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,768,12288,0.011250666446155973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,12288,0.014183999763594734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,768,10240,0.010388444695207808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,768,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,768,51200,0.024738666084077623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,65536,0.056792888376447886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,768,8192,0.009000000026490953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,10240,0.01276711126168569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,8192,0.011417778001891242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,1536,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,12288,0.013799111048380533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,7168,0.010061333576838175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,768,6144,0.009714666340086196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,7168,0.010713777608341642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,10240,0.01221422188811832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,6144,0.009890666438473595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,6144,0.009364444348547194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,5120,0.009018667042255402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,16384,0.017044444878896076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,768,4096,0.008051555189821456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,4096,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,768,3584,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,4096,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,3584,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,768,3072,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,3584,0.007667555577225155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,3072,0.007696888513035244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,768,2560,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,51200,0.04297599858707852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,768,2048,0.0053093334039052325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,2560,0.007602666815121968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,2560,0.006813333266311222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,768,1536,0.004568888909286923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,2048,0.006485333459244833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,1536,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,768,1024,0.0042133331298828125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,3072,0.007080000307824876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,768,7168,0.01037333326207267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,1024,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,1024,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,2048,0.007144889069928064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,768,0.005548444473081165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,1536,0.006953777538405524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,512,0.0052826665341854095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,768,256,0.003195555466744635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,8192,0.01070666644308302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,768,768,0.0035537779331207275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,768,128,0.0028684443483750024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,256,0.005007110950019625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,768,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,128,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,128,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,768,32,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,768,5120,0.009180444810125563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,64,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,512,0.005977777971161737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,32,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,5120,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,65536,0.05262577864858839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,256,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,512,16384,0.013077333569526672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,51200,0.04246222310596042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,16384,0.016844444804721408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,16384,0.016568000117937725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,512,12288,0.010952889091438718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,512,65536,0.0236462222205268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,512,10240,0.010072888599501716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,12288,0.014164444473054675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,12288,0.013776889277829064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,512,51200,0.01973777843846215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,512,8192,0.00907466643386417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,65536,0.05625066492292616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,10240,0.01257777793539895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,512,7168,0.008393777741326226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,8192,0.011414222419261932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,51200,0.042062222957611084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,512,6144,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,8192,0.010508444574144153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,512,5120,0.008674666285514832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,7168,0.010836444795131683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,6144,0.010057777994208867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,512,4096,0.0076657773719893555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,6144,0.009369778136412302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,5120,0.009038222332795462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,5120,0.008606221940782335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,4096,0.00869688888390859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,4096,0.007924444145626491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,512,3072,0.006592888798978593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,10240,0.012808889150619507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,3584,0.0073626670572492815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,512,2560,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,3072,0.008001777860853408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,3072,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,768,512,0.0034640000926123727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,512,2048,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,2560,0.007650666766696506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,2560,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,2048,0.007289778027269576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,7168,0.010198222266303169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,2048,0.00648266656531228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,512,3584,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,1536,0.005987555616431766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,512,768,0.0038720000949170855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,1024,0.0064382221963670515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,1024,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,768,0.006265777680608962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,768,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,512,512,0.0035537779331207275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,512,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,512,256,0.003208000005947219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,512,0.0052791109515561005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,512,1024,0.004212444441186057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,1536,0.0069164443347189165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,512,128,0.002959111084540685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,256,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,3584,0.008017777568764156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,128,0.006050666587220297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,512,32,0.0026551110463009942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,768,64,0.0025315555847353404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,256,65536,0.019703111714786954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,512,1536,0.004586666822433472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,32,0.005016889009210798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,65536,0.049633777803844877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,65536,0.051683555046717324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,256,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,256,16384,0.01015822258260515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,512,64,0.0028853333658642242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,128,0.004935111022657818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,256,12288,0.012411555482281579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,16384,0.016190222567982145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,64,0.004910222358173794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,256,10240,0.011394666300879585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,12288,0.014073777529928418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,12288,0.013566222455766467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,256,8192,0.010029333333174387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,10240,0.012776888906955719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,256,51200,0.017313778400421143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,256,7168,0.009402666654851701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,8192,0.010432889064153036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,51200,0.03964444332652622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,51200,0.04167022307713827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,7168,0.010781333678298526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,16384,0.016536888149049547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,256,5120,0.008447999755541483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,6144,0.009988444546858469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,7168,0.010138666464222802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,256,4096,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,5120,0.009188444250159794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,4096,0.00868800034125646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,256,3584,0.007048889166778988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,10240,0.011866666376590729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,256,3072,0.006321777900060018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,3584,0.008057778080304464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,256,6144,0.008664000365469191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,256,2560,0.005592888842026393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,3072,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,2560,0.007633777956167857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,2560,0.006649777707126405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,5120,0.008625778059164682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,2048,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,2048,0.006313777632183499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,256,1536,0.004280888785918553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,8192,0.011334222224023608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,3584,0.0075937774446275495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,1536,0.006670222100284364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,3072,0.008001777860853408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,256,768,0.003591999825504091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,1024,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,1024,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,256,512,0.0032275555034478507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,6144,0.009114666945404476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,768,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,256,2048,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,256,256,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,512,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,256,128,0.0028808888875775864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,256,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,512,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,256,64,0.00281866660548581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,128,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,256,32,0.002519999941190084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,128,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,1536,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,128,65536,0.01585422290696038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,64,0.00498488876554701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,128,51200,0.013691555294725629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,65536,0.04873866505093045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,768,0.005316444569163852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,128,16384,0.009753777749008602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,51200,0.03943822118971083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,51200,0.039629333549075656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,256,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,128,12288,0.009692444569534725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,256,1024,0.0038746665749284956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,128,10240,0.009024000002278222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,12288,0.014118222726715935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,12288,0.013468444347381592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,10240,0.012778666284349231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,10240,0.01183111137813992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,32,0.004976888911591636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,65536,0.04963555600908068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,128,7168,0.00903555585278405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,8192,0.010402667025725046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,128,6144,0.008672888908121321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,7168,0.009933333430025313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,16384,0.01658933361371358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,16384,0.01660711069901784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,128,5120,0.008656889200210571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,6144,0.009713778065310584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,6144,0.009000000026490953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,128,4096,0.007634667058785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,128,8192,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,5120,0.008554666406578487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,128,3584,0.0069546666410234236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,8192,0.011197333534558615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,4096,0.007641777396202087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,128,3072,0.0063075555695427795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,7168,0.010553778045707278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,4096,0.007829333345095316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,128,2560,0.005618666609128316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,3072,0.0069546666410234236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,2560,0.007661333514584436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,2560,0.006664888726340399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,5120,0.00903200027015474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,4096,0.008374222450786168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,128,1536,0.004562666846646203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,2048,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,3584,0.007991111112965478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,3584,0.007383999725182851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,3072,0.007670222057236566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,1024,0.0063279999627007385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,128,2048,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,768,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,128,512,0.003212444484233856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,128,1024,0.003935111065705617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,512,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,128,256,0.002862222285734283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,1024,0.005670222143332164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,128,768,0.003552888830502828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,128,128,0.0028799999919202593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,256,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,2048,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,768,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,128,0.005973333285914526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,128,0.004968889057636261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,64,0.005094222310516569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,64,65536,0.014852444330851236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,32,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,64,51200,0.014138667119873894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,65536,0.049136890305413135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,64,16384,0.009656888743241629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,256,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,51200,0.039614223771625094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,128,32,0.0028444443725877335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,16384,0.016202666693263583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,64,10240,0.00904088881280687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,64,8192,0.008035555481910706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,10240,0.011741333537631564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,64,7168,0.009000000026490953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,8192,0.010430222584141625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,64,6144,0.008375110725561777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,128,64,0.002532444480392668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,64,5120,0.007678222325113084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,512,0.0052897776994440295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,64,4096,0.0069413334131240845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,5120,0.008344000412358178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,64,3584,0.006646222124497096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,4096,0.007620444728268518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,64,3072,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,64,12288,0.009684444301658207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,64,2560,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,3072,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,64,2048,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,1536,0.0070266665683852295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,64,1536,0.004254222330119875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,2048,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,64,1024,0.0038684445122877755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,1536,0.005980444451173146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,64,768,0.003576888806290097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,3584,0.007275555696752336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,64,512,0.0032088889016045462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,1024,0.005590222362014983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,64,256,0.003027555636233754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,768,0.005471999860472149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,512,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,1536,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,256,0.005046222358942032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,128,0.004942222187916438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,64,32,0.00253866674999396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,64,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,32,65536,0.0281333327293396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,7168,0.009718221922715506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,32,51200,0.024733333124054804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,65536,0.04899911085764567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,64,128,0.0028844444702068963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,12288,0.01344622257683012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,64,64,0.002526222210791376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,51200,0.039431111680136785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,32,10240,0.008672888908121321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,16384,0.01625066664483812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,32,8192,0.008027555214034187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,12288,0.013439999686347114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,32,7168,0.009011555876996782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,10240,0.011748444702890186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,32,0.004942222187916438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,8192,0.010454222559928894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,7168,0.009934222532643212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,32,5120,0.007639110916190677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,2560,0.006568000134494569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,6144,0.009388444324334463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,32,4096,0.007300444775157505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,32,3584,0.006637333167923822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,4096,0.007688889073001013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,32,3072,0.006512889017661412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,3584,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,32,2560,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,32,6144,0.008345777789751688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,32,2048,0.005619555711746216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,3072,0.00702311098575592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,32,1536,0.0042568888101312844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,32,12288,0.009655999640623728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,32,1024,0.003900444342030419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,5120,0.00833155545923445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,32,768,0.0035457776652442086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,32,512,0.00319377767542998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,1024,0.00564177789621883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,32,256,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,768,0.0053324442770746015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,32,128,0.0028204443968004654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,512,0.005280888742870755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,32,64,0.0028515555378463534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,256,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,32,32,0.0028684443483750024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,128,0.004947555561860402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,64,0.004946666873163647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,32,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,6144,0.009374221993817223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,129,32,16384,0.009345778160625035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,2560,0.006575111299753189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,51200,1.1428391138712566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,2048,0.006244444598754247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,16384,0.3750844531589084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,16384,0.20704621738857695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,12288,0.27946845690409344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,16384,0.1365448898739285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,12288,0.15165066719055176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,51200,0.3995199998219808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,12288,0.10353777805964152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,10240,0.23682755894131133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,1536,0.005982222242487802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,10240,0.1349875529607137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,10240,0.0853475530942281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,51200,0.6206871138678657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,8192,0.11116177505917019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,16384,0.30044621891445583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,12288,0.23448533482021758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,7168,0.17109244399600557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,10240,0.19361066818237305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,8192,0.06995377937952678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,7168,0.09529866774876912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,8192,0.15247110525767008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,7168,0.062315556738111705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,6144,0.07980977826648288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,5120,0.12320889366997613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,6144,0.05438489052984449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,7168,0.13908444510565862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,5120,0.06702844301859538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,5120,0.046563555796941124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,6144,0.14651555485195583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,5120,0.0980000032318963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,8192,0.19624355104234484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,3584,0.08883644474877252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,4096,0.0795653329955207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,3584,0.049754665957556836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,3072,0.07807822359932794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,3584,0.07160622543758817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,51200,0.9283146328396268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,3072,0.04338577720854017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,2560,0.06551200151443481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,4096,0.03964266512129042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,3584,0.035644445154401995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,6144,0.11702666017744277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,2560,0.028064890040291682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,2048,0.052341332038243614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,4096,0.05514666769239637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,2048,0.03266311022970412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,3072,0.06367466847101848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,3072,0.04094311263826158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,2048,0.024343111448817786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,1536,0.04083289040459527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,2048,0.04718933171696133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,2560,0.03773244553142124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,1536,0.01943111088540819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,4096,0.09967288706037734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,768,0.022625777456495497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,1024,0.019840000404251944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,1024,0.03164533442921109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,768,0.016195555528004963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,768,0.014230221509933472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,512,0.015843555331230164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,1536,0.02585244509908888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,512,0.023638221952650283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,256,0.010646222366227044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,512,0.01330133361948861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,2560,0.055171555942959256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,1024,0.01588622232278188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,256,0.011771555576059552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,128,0.00832799987660514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,256,0.019150222341219585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,768,0.02842666705449422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,128,0.011927111281288994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,128,0.011570666399266986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,1024,0.0303164455625746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,128,0.016732444365819294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,32,0.006337777607970768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,1536,0.0399582220448388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,64,0.011447110937701331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,32,0.014836443795098199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,256,0.012404444317022959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,64,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,512,0.013607111242082385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,65536,1.116349326239692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,65536,0.6357235378689235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,51200,0.8807119793362088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,16384,0.2828755643632677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,51200,0.48805422253078884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,16384,0.16467821598052979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,16384,0.10974755552079941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,12288,0.21699733204311797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,12288,0.12504177623324925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,51200,0.32691823111640084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,65536,0.4107235537634955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,10240,0.16782666577233207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,16384,0.24077510833740234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,10240,0.10529777738783096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,10240,0.06980178091261122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,12288,0.1842577722337511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,8192,0.08634399705462986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,8192,0.0563102232085334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,10240,0.15671911504533556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,12288,0.08180266618728638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,8192,0.1285582251018948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,7168,0.05043111244837443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,6144,0.10376177893744574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,8192,0.13849688900841609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,7168,0.11162932713826497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,6144,0.06428177489174737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,5120,0.0901671118206448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,7168,0.12024533748626709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,6144,0.09818133380677964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,5120,0.053929779264662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,4096,0.07414488659964667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,5120,0.03786755601565043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,5120,0.08147999975416395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,51200,0.7641510963439941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,3584,0.06478577852249146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,4096,0.033129778173234724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,4096,0.06808355781767103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,6144,0.04623022344377306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,3584,0.04189777705404493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,65536,0.9917066362169055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,3584,0.06037066380182902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,2560,0.04782844583193461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,3072,0.035465776920318604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,3072,0.033187554942237005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,7168,0.07509777943293254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,3072,0.054698665936787925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,2560,0.031003554662068684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,2560,0.0244186669588089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,2048,0.039260443713929914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,1536,0.030341333813137476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,2560,0.048264887597825795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,2048,0.021012443635198805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,2048,0.04066133168008592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,3584,0.029265777932273015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,1024,0.02388711108101739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,1536,0.021525333325068157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,1536,0.03478666808870103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,768,0.01719111038578881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,1024,0.016418667303191293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,1024,0.02774133284886678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,768,0.013464888764752282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,768,0.012823111481136747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,4096,0.04477599925465054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,2048,0.02716444432735443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,512,0.011722666521867117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,512,0.011805333197116852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,256,0.009043555292818282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,1024,0.013949332965744866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,256,0.01074044406414032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,128,0.006601777755551868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,256,0.010391111175219217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,256,0.017048888736300997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,512,0.012308444413873883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,768,0.023911111884646948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,128,0.009872888525327047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,32,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,3072,0.05558400021659004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,1536,0.01685333251953125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,64,0.009920888476901585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,512,0.020786666207843356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,32,0.012777778009573618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,65536,0.3223377863566081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,51200,0.2530426714155409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,65536,0.14040799935658774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,128,0.01075644459989336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,51200,0.1757964425616794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,128,0.014691554837756686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,16384,0.08680355548858643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,51200,0.113155550426907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,16384,0.061549332406785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,12288,0.0654204421573215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,65536,0.2301022211710612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,16384,0.04103199972046746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,16384,0.05909599860509237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,12288,0.046738667620552905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,10240,0.05660711394415962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,12288,0.03237777617242601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,12288,0.0444106658299764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,10240,0.027951998843087092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,8192,0.04634311132960849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,10240,0.03909600112173293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,8192,0.023687111006842718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,7168,0.0410773323641883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,8192,0.03401866555213928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,64,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,65536,0.2008977731068929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,7168,0.030015998416476782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,7168,0.02197866638501485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,51200,0.15853688451978895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,5120,0.030755556292004053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,6144,0.02677511175473531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,6144,0.019331556227472093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,8192,0.034479111433029175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,5120,0.017090666625234816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,4096,0.02644444505373637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,5120,0.023966223001480103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,4096,0.015108444624476962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,6144,0.03531644410557217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,3584,0.023010666171709698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,7168,0.029822223716311987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,3584,0.01756000022093455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,3584,0.014677332507239448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,3072,0.020245333512624104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,3584,0.01921955578856998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,6144,0.027611555324660406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,2560,0.01754133403301239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,3072,0.017197334104114108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,3072,0.015836444165971544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,4096,0.019623999794324238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,4096,0.0210524449745814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,2048,0.014480888843536377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,10240,0.040072888135910034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,2560,0.015832889411184523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,1536,0.011386666860845355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,2048,0.011240888800885944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,2048,0.014864888456132678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,5120,0.023021333747439917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,1536,0.009530666801664565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,1536,0.013462222284740873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,2560,0.014056889547242058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,1024,0.009039999710188972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,2560,0.012504888905419244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,768,0.007672888537247975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,1024,0.011956444217099084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,768,0.007978666987684038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,768,0.007743111087216272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,512,0.006396444307433234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,768,0.010479110810491772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,3072,0.01312355531586541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,1024,0.009510222408506606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,512,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,512,0.009016888837019602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,256,0.006605333338181178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,128,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,256,0.008357333640257517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,2048,0.012433778080675336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,128,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,64,0.003573333223660787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,128,0.007028444773621029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,1536,0.010738666686746808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,32,0.004218666503826777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,256,0.00434044458799892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,64,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,65536,0.2471840116712782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,32,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,256,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,65536,0.1782888836330838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,51200,0.18819199668036568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,65536,0.12811377313401964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,51200,0.1358799934387207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,16384,0.06665777497821383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,51200,0.10114133358001709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,16384,0.04843466811709934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,12288,0.05026222268740336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,16384,0.03780711028310988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,1024,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,12288,0.03700266612900628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,12288,0.030001779397328694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,128,0.008393777741326226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,512,0.007366221812036302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,10240,0.032159109910329185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,8192,0.03623200125164456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,10240,0.025801777839660645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,16384,0.04963733421431648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,65536,0.17071821954515246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,8192,0.027781334188249376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,12288,0.038776000340779625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,8192,0.022101332743962605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,7168,0.024317334095637005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,7168,0.026664889521068994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,10240,0.034240000777774386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,7168,0.01998933321899838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,51200,0.1360728873146905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,6144,0.02164977788925171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,6144,0.01812355551454756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,5120,0.023764444722069636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,5120,0.01849333279662662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,5120,0.021356445219781663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,4096,0.020791111720932853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,6144,0.027424888478385076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,4096,0.014819555812411837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,4096,0.01590133375591702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,3584,0.01753777762254079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,4096,0.018904889623324077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,6144,0.024146666129430134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,3072,0.015672889020707872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,3584,0.013898667362001208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,3584,0.014615111880832247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,3584,0.01739022301303016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,2560,0.013398222625255585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,3072,0.012649777862760754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,3072,0.01312355531586541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,5120,0.0162462227874332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,2048,0.011672889192899069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,2560,0.011924444801277585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,2560,0.012085333466529846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,2560,0.015388444066047668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,2048,0.01038044442733129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,1536,0.009716444545321995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,10240,0.042709334029091724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,1536,0.009617778162161509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,1024,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,1536,0.012994666894276937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,3072,0.015992888145976596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,1024,0.00740533322095871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,768,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,1024,0.010808000134097205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,2048,0.010794666906197866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,8192,0.02981511089536879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,768,0.00685511115524504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,512,0.004888888862397936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,768,0.007526222202512953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,768,0.009745777481132084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,1536,0.008756444685988957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,256,0.0042133331298828125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,512,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,512,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,128,0.003899555653333664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,256,0.0069511110583941145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,256,0.0061048890153567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,256,0.008615111311276754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,1024,0.008257778154479133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,64,0.0038693332009845306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,128,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,128,0.008340444829728868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,7168,0.031122667921913996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,64,0.006113777971929974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,32,0.006959999601046245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,65536,0.20483022265964082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,51200,0.15745333830515543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,512,0.009407111340098912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,65536,0.12064977486928304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,128,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,51200,0.13023377789391413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,16384,0.05512533254093594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,51200,0.09661600324842666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,2048,0.013552889227867126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,12288,0.04253510965241326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,16384,0.036102222071753606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,16384,0.04733777708477444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,51200,0.12144088745117188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,10240,0.03616266780429416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,12288,0.028023110495673284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,12288,0.037267555793126426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,32,0.003927999900446998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,8192,0.03018666638268365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,10240,0.03249333302179972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,65536,0.15145066049363878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,8192,0.026519111461109583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,8192,0.028232001596026953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,65536,0.16707111729515922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,7168,0.023136888941129048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,10240,0.024728000164031982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,6144,0.023376888699001733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,7168,0.0258879992696974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,16384,0.04636355572276645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,6144,0.020808888806237113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,5120,0.02026844521363576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,12288,0.035641776190863714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,5120,0.017888888716697693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,10240,0.03085244364208645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,5120,0.020534222324689228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,5120,0.015927111109097798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,4096,0.015496888094478183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,6144,0.023763555619451735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,4096,0.014138667119873894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,4096,0.018215111560291715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,3584,0.014906666345066495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,4096,0.01723288827472263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,7168,0.01908266709910499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,3584,0.014135110709402295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,3584,0.017324444320466783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,3072,0.01311377767059538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,6144,0.017655111021465726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,2560,0.011725333001878528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,3072,0.01605599953068627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,3072,0.01239733315176434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,2048,0.010047111246320937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,2560,0.011741333537631564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,2560,0.01514488955338796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,3072,0.013433777623706393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,2048,0.010659555594126383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,2048,0.014077333940400017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,7168,0.0265964451763365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,1536,0.01276000009642707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,1024,0.006895111252864202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,2560,0.011056889262464313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,1024,0.007854222423500484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,3584,0.013210666676362356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,1024,0.01015911085738076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,1024,0.007670222057236566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,8192,0.021406221720907424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,768,0.006637333167923822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,512,0.004497777670621872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,1536,0.008424888882372115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,512,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,512,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,1536,0.008658666577604082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,2048,0.009729777773221334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,768,0.005625777774386936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,256,0.006788444601827198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,1536,0.008661333057615492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,128,0.003573333223660787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,256,0.008358221915033128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,768,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,64,0.0035457776652442086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,128,0.006799110935793982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,32,0.003539555602603488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,64,0.00599644457300504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,256,0.0039022221333450745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,32,0.006618666566080517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,65536,0.1801555554072062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,256,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,65536,0.12639466921488443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,768,0.009687110781669617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,65536,0.07168266508314344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,128,0.005970666805903117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,128,0.00850133349498113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,16384,0.04469244347678291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,51200,0.098380446434021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,65536,0.14029244581858316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,16384,0.03507644600338406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,16384,0.023338667220539514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,16384,0.04299111167589823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,51200,0.11322666539086236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,10240,0.03018222252527873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,12288,0.01903555625014835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,51200,0.1458817720413208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,12288,0.03489777776930068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,10240,0.02401688860522376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,8192,0.025578666064474318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,10240,0.0166320006052653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,10240,0.03035466538535224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,51200,0.05832533041636149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,7168,0.02257333364751604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,512,0.006960888703664144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,8192,0.02100177771515316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,12288,0.035939554373423256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,6144,0.019860444797409903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,8192,0.02628444466325972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,7168,0.013789333403110504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,7168,0.01794399983353085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,5120,0.016963554753197562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,7168,0.023958222733603582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,12288,0.027664000789324444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,4096,0.014349333114094205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,5120,0.011748444702890186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,5120,0.014318222800890604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,5120,0.01964000033007728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,3584,0.012682666381200155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,4096,0.011037333144081963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,4096,0.012797333300113678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,8192,0.014783999986118741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,3072,0.011766222616036734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,3584,0.009773333039548662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,3584,0.01201600001917945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,3584,0.016188444362746347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,2560,0.010024888647927178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,3072,0.011002666420406766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,6144,0.01640533407529195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,3072,0.01552977826860216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,6144,0.022026666336589392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,2048,0.009000000026490953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,2560,0.00869333330127928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,2560,0.014144000079896716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,2048,0.00905777762333552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,1536,0.007847111258241866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,2048,0.01348888874053955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,1536,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,3072,0.00902133352226681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,6144,0.012871111432711283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,1024,0.006261333409282897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,1536,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,2560,0.010090666512648264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,1024,0.006681777950790193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,1024,0.009703999592198266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,4096,0.01745244363943736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,768,0.005998222364319696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,768,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,768,0.008671999805503422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,1536,0.011756444142924415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,512,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,256,0.0038968887594011095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,512,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,1024,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,768,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,128,0.003551111039188173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,256,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,256,0.005626666463083691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,2048,0.007717333734035492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,256,0.007705777883529663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,128,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,512,0.004261333495378494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,128,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,128,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,64,0.005646222167544895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,512,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,32,0.006189333481921091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,65536,0.12574399842156306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,51200,0.12774488661024305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,65536,0.06936711072921753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,16384,0.04506577716933357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,64,0.0033813334173626373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,65536,0.13887643814086914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,51200,0.05637600024541219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,32,0.003584889074166616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,12288,0.03609600000911289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,16384,0.022083555658658344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,65536,0.16253244876861572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,16384,0.0426879988776313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,12288,0.018231999542978074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,12288,0.034154666794670954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,10240,0.02309600015481313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,10240,0.015967110792795818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,10240,0.029822223716311987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,51200,0.09551644325256348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,8192,0.02036088870631324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,8192,0.01426755554146237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,16384,0.034719109535217285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,51200,0.11020444499121772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,10240,0.02993688980738322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,7168,0.013388444152143268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,7168,0.02370133333735996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,8192,0.02626133296224806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,6144,0.016068443655967712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,5120,0.016971555021074083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,6144,0.021706667211320665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,7168,0.022505778405401442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,5120,0.014119111829333834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,4096,0.014509333504570855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,12288,0.026885333988401625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,6144,0.019634667370054457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,4096,0.010273777776294285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,4096,0.012703110774358114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,3584,0.012210666305489011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,6144,0.012353777885437012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,3584,0.011786667009194693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,8192,0.02589599953757392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,3584,0.01625066664483812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,5120,0.011413333316644033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,7168,0.017551110850440133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,3072,0.010808000134097205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,2560,0.009669333696365356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,3072,0.0086666668454806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,3072,0.015241778559154935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,2048,0.0086986662613021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,2560,0.009871111147933537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,2560,0.014127110441525778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,3584,0.00907466643386417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,2048,0.007327111230956183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,1536,0.007516444557242923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,2048,0.013213333156373767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,5120,0.019861333900027804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,1536,0.00702044450574451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,1536,0.0073831114504072405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,1536,0.010959111154079437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,4096,0.01724444495307075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,1024,0.006643555644485686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,768,0.004954666727119022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,1024,0.009714666340086196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,768,0.00639111093348927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,2048,0.008637333081828224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,512,0.004565333326657613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,768,0.0090515555606948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,1024,0.005912888795137405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,512,0.00867377801073922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,512,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,256,0.003863999827040566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,2560,0.008100444244013892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,256,0.006112888869312074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,256,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,128,0.003547555456558863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,256,0.008247111406591203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,3072,0.011050666371981302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,64,0.0035386664999855887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,128,0.005983110931184556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,128,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,32,0.003569777641031477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,512,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,64,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,32,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,65536,0.11696267127990723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,65536,0.06679466697904798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,51200,0.11584356096055771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,51200,0.09033066696590847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,51200,0.05426222085952759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,65536,0.14669599797990587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,51200,0.11053599913914998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,1024,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,16384,0.033506668276256986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,12288,0.029024001624849107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,16384,0.021406221720907424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,16384,0.04245511028501722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,768,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,12288,0.01754133403301239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,12288,0.03376266691419814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,65536,0.13830755816565618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,8192,0.02109777761830224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,10240,0.015642666154437594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,10240,0.029951112137900457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,8192,0.01402311192618476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,8192,0.019295111298561096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,7168,0.018186666899257235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,8192,0.025237333443429735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,7168,0.013096888860066732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,10240,0.024358222881952923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,7168,0.023373333944214716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,16384,0.036333332459131874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,6144,0.01551555593808492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,5120,0.013810666898886362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,6144,0.021436444587177698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,5120,0.013857777747843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,5120,0.019497777024904888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,12288,0.025780444343884785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,128,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,4096,0.012461333639091916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,6144,0.012087999946541257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,4096,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,3584,0.010844444235165914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,4096,0.016872000363137987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,3584,0.009004444711738164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,3072,0.009869332942697737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,7168,0.017123555143674214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,3584,0.016240888171725802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,3072,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,3072,0.01072266697883606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,2560,0.009062222308582729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,3072,0.01475288967291514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,2560,0.009346666435400644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,2048,0.008047111332416534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,3584,0.011476444701353708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,4096,0.01254488858911726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,2048,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,1536,0.006780444333950679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,2048,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,10240,0.022322666313913133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,1536,0.0070159998204973005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,1024,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,2560,0.008015111088752747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,2560,0.014095111025704278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,1024,0.006676444576846228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,6144,0.015819556183285184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,1024,0.009354666703277165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,2048,0.012102222277058495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,768,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,512,0.0038782221575578055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,1536,0.01109866632355584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,512,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,1024,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,512,0.00869955536392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,512,0.00630044440428416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,256,0.005616888817813661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,256,0.008287111090289222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,768,0.009012444151772393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,128,0.003551111039188173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,256,0.006007110906971826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,128,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,128,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,128,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,32,0.003495111233658261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,256,0.0035128887328836652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,64,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,65536,0.12280889352162679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,32,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,768,0.0063564446237352155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,65536,0.11091199848386978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,51200,0.09640355242623223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,65536,0.06431021955278185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,64,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,65536,0.13855555322435167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,16384,0.039841777748531766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,51200,0.0865280032157898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,51200,0.10970311032401191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,16384,0.031672000885009766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,1536,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,12288,0.0313351121213701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,16384,0.04177511069509718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,12288,0.017071111334694754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,10240,0.02224622170130412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,768,0.004553777890072929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,51200,0.05253155695067512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,10240,0.0151555554734336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,8192,0.019066666563351948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,16384,0.020615999897321064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,8192,0.018770666586028207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,7168,0.016511999898486666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,8192,0.01349244432316886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,12288,0.033245331711239286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,7168,0.012588444683286877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,6144,0.01482399966981676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,10240,0.021294222937689886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,7168,0.02310933338271247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,10240,0.02948533495267232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,5120,0.012880889077981314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,6144,0.011788444386588203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,6144,0.01515111161602868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,6144,0.021073778470357258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,4096,0.01144088887506061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,12288,0.024481778343518574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,5120,0.01348355578051673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,8192,0.02494577732351091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,3584,0.010052444206343757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,4096,0.009093333449628618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,4096,0.012150222228633033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,7168,0.016707556115256417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,3072,0.009039999710188972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,3584,0.011353777514563667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,3584,0.015832889411184523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,3072,0.009757333331637913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,2560,0.0083155557513237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,3072,0.014577777849303352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,5120,0.01036800030204985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,5120,0.019069333871205647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,2048,0.007275555696752336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,2560,0.007768889268239339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,2560,0.01349866638580958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,2048,0.007970666719807519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,2048,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,2048,0.01184977756606208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,1536,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,1024,0.004574222283230888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,3072,0.008394666843944127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,1536,0.010728889041476779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,1024,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,1024,0.0063493334584765965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,768,0.00423200014564726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,2560,0.00870400004916721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,4096,0.01662400033738878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,1536,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,3584,0.008685333033402761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,768,0.009000889129108852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,512,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,1536,0.00703288863102595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,256,0.0035537779331207275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,512,0.008348444269763099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,256,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,256,0.00562755556570159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,128,0.003531555748648114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,256,0.007984889050324758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,512,0.0038977778620190094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,128,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,64,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,128,0.00535999983549118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,128,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,512,0.005592888842026393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,64,0.005611555443869696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,32,0.005984888722499211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,65536,0.07014133532842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,1024,0.009705777797434065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,51200,0.07884799771838717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,768,0.006292444550328785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,32,0.0035377778112888336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,16384,0.03196888830926683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,51200,0.054457777076297335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,51200,0.050168000989490084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,65536,0.09472444322374131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,12288,0.025783111651738484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,16384,0.019880000087949965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,16384,0.02109333376089732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,16384,0.041414221127827965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,12288,0.016534222496880423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,65536,0.13797956042819554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,768,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,10240,0.014831999937693277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,10240,0.01484266585773892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,10240,0.029084444046020508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,65536,0.06212800078921848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,10240,0.019179556104871962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,12288,0.033054222663243614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,7168,0.014461333552996317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,8192,0.02475999957985348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,7168,0.01202666676706738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,8192,0.016146666473812528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,6144,0.012789333032237159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,7168,0.012087110843923358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,8192,0.013442666994200813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,6144,0.01108000013563368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,5120,0.0114702218108707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,6144,0.020741333564122517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,8192,0.013100444442696042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,5120,0.009736888938479954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,4096,0.010471111370457543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,5120,0.018952889574898612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,51200,0.10929599735471938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,7168,0.02275644408331977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,3584,0.009351111120647853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,4096,0.008681777450773451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,4096,0.016578666037983365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,3584,0.008445333275530074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,3584,0.00851644410027398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,6144,0.011027555498811932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,3072,0.007763555480374231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,3072,0.008023999631404877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,2560,0.007859555383523306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,5120,0.010110222631030613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,2560,0.007672888537247975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,2560,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,4096,0.009147555463843876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,12288,0.016869333055284288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,3072,0.00831911133395301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,2048,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,1536,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,3584,0.015517334143320719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,2048,0.011736888852384357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,1024,0.0042257776690853965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,1536,0.006991110742092133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,1536,0.006163555714819167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,3072,0.014340443743599786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,1536,0.010633777413103314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,1024,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,1024,0.005716444303592046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,2048,0.006590222318967183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,2560,0.012784000072214337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,768,0.006331555545330048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,768,0.009027555584907532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,512,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,256,0.003237333355678452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,512,0.008653333617581261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,768,0.0039013334446483185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,256,0.005116444494989184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,128,0.0032177778581778207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,1024,0.009711999860074785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,768,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,128,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,64,0.00319377767542998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,128,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,32,0.0035422220826148987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,64,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,32,0.005564444594913059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,256,0.008020444048775567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,512,0.0035822221802340615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,65536,0.06766044431262545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,128,0.00500266667869356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,51200,0.07003377543555366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,2048,0.007324444750944774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,65536,0.1380231115553114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,16384,0.029532445801628962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,51200,0.053195556004842125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,51200,0.049172444476021655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,51200,0.1098640031284756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,16384,0.019630221856964957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,12288,0.024705777565638225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,16384,0.020408888657887776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,16384,0.04109244545300802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,10240,0.02165777815712823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,12288,0.016338666280110676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,12288,0.016344888342751395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,12288,0.03275022241804335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,8192,0.015447111593352424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,10240,0.014476444986131458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,10240,0.014231999715169271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,10240,0.028858668274349634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,8192,0.01276000009642707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,7168,0.013778666655222574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,65536,0.0844328867064582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,256,0.005967999911970562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,6144,0.012181333369678922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,7168,0.011568889021873474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,7168,0.022777777579095628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,65536,0.06098844607671102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,6144,0.01107377807299296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,512,0.006177777631415262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,6144,0.020625778370433383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,5120,0.010149333212110732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,8192,0.024682665864626568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,5120,0.009406222237481011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,7168,0.011766222616036734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,4096,0.008722667064931657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,6144,0.010690666735172272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,5120,0.010989333192507425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,3584,0.008887111312813228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,4096,0.008707555631796518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,3584,0.008375999828179678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,8192,0.01312711089849472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,5120,0.018960000740157235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,3584,0.015406222807036506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,3072,0.007303111255168915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,2560,0.007531555162535773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,3072,0.013791999883121915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,4096,0.016187555260128446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,2560,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,2048,0.0063884444534778595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,2560,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,2560,0.012787555654843649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,3584,0.007676444119877285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,1536,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,4096,0.009661333428488838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,2048,0.011063111325105032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,1024,0.004229333251714706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,1536,0.010159999959998662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,1024,0.005591111050711737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,1024,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,768,0.0038817777401871155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,3072,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,2048,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,2048,0.006364444477690591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,512,0.003662222375472387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,768,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,3072,0.008011555506123437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,768,0.00904888908068339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,256,0.0035706667436493766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,512,0.005987555616431766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,512,0.005290666802061929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,512,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,256,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,256,0.005247999810510212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,256,0.00775822252035141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,128,0.005025777965784073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,64,0.003208000005947219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,768,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,1536,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,64,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,32,0.005374222166008419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,128,0.00318755561278926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,1024,0.006642666541867786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,65536,0.06605688730875652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,128,0.005976000179847081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,128,0.008030222521887885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,65536,0.13779200447930232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,1536,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,16384,0.02548800077703264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,51200,0.051316443416807384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,51200,0.04789244466357761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,65536,0.07754044400321113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,16384,0.019300444258583915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,16384,0.04083911246723599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,65536,0.05977955791685316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,12288,0.015836444165971544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,51200,0.06351821952395968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,12288,0.01611911091539595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,51200,0.10966844691170587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,10240,0.014152000347773233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,16384,0.020089776979552377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,10240,0.028551111618677776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,10240,0.014281777871979607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,8192,0.012345777617560493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,10240,0.018184888694021437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,7168,0.013424889081054263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,32,0.0031902222997612427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,7168,0.011095999843544431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,6144,0.012124444047609964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,7168,0.022640888889630634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,8192,0.016516443755891588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,6144,0.010841777755154504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,8192,0.0124177775449223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,6144,0.020608888732062448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,8192,0.02456444501876831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,12288,0.032637334532207914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,5120,0.009696889254781935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,4096,0.009434666898515489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,5120,0.018943111101786297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,7168,0.011454222102959951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,3584,0.008608889248636033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,4096,0.008676444490750631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,4096,0.008008889026112026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,4096,0.01586844523747762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,3584,0.008300444318188561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,3584,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,5120,0.01053244454993142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,3584,0.014736889137162102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,2560,0.007124444676770105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,3072,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,3072,0.013474666410022311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,5120,0.009702222214804756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,2560,0.006763555523422029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,2048,0.006244444598754247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,12288,0.021495110458797876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,2560,0.012112889024946423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,3072,0.007975999679830339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,2048,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,2048,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,2048,0.011391999820868174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,3072,0.007343110938866933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,1536,0.006089777582221561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,1536,0.010381333529949188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,1024,0.005658666706747479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,1024,0.009009777671760982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,768,0.004267555558019214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,6144,0.010232888989978367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,1536,0.004943111290534337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,768,0.005555555638339784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,1536,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,1024,0.004570666700601578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,512,0.0052826665341854095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,2560,0.007672000262472365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,512,0.008358221915033128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,256,0.003556444413132138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,512,0.00601333338353369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,1024,0.00654666663871871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,128,0.003253333270549774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,256,0.005224888937340842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,256,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,768,0.0063475556671619415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,512,0.0038720000949170855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,128,0.004955555415815777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,128,0.00795377790927887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,65536,0.064337776766883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,64,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,32,0.0053022222386466126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,65536,0.06260266568925646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,256,0.005977777971161737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,51200,0.05349777804480659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,65536,0.1377964417139689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,64,0.0032035555276605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,128,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,768,0.009023110899660323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,51200,0.049172444476021655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,51200,0.10925866497887506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,16384,0.019253333409627277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,16384,0.019304000669055514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,16384,0.040468444426854454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,65536,0.05856621927685208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,12288,0.015501333607567681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,10240,0.017321777012613084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,12288,0.032319999403423734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,10240,0.013863999810483722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,10240,0.013781333135233985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,8192,0.015326221783955893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,10240,0.028535998529858057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,12288,0.01977155605951945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,8192,0.011782222323947482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,8192,0.012069332930776807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,8192,0.024487111303541396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,12288,0.015499555402331881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,7168,0.010732444624106089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,7168,0.011457777685589261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,7168,0.02199199961291419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,16384,0.022661333282788593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,32,0.003212444484233856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,5120,0.010427555276287926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,6144,0.020651555723614164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,5120,0.009324444664849175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,7168,0.013849777479966482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,4096,0.009153777526484596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,5120,0.018271999226676095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,4096,0.00850399997499254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,4096,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,3584,0.008037333687146505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,4096,0.015439111325475903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,6144,0.010079111489984725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,3584,0.007661333514584436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,3584,0.014361777239375643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,5120,0.009352000223265754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,3072,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,2560,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,3072,0.013404444687896304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,6144,0.013103110922707452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,2560,0.006979555719428592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,2560,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,3584,0.008347555167145198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,51200,0.0466195543607076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,2048,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,1536,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,2048,0.011235555840863122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,3072,0.007711110843552484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,1536,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,1536,0.0069822221994400024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,2048,0.005591111050711737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,6144,0.010371555884679159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,3072,0.007640000018808577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,1024,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,768,0.004226666771703296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,1024,0.006592000110281839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,2048,0.007183999651008182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,1024,0.009425777528021071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,512,0.0039022221333450745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,768,0.005415110952324337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,768,0.009002666506502364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,512,0.005280888742870755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,1536,0.01037688884470198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,256,0.003212444484233856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,2560,0.01239911135700014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,256,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,128,0.0032115555885765287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,768,0.0063271112740039825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,128,0.004935111022657818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,1024,0.004243555582231945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,512,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,128,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,32,0.002962666667169995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,256,0.004952888935804367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,64,0.004941333499219683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,256,0.00796622203456031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,65536,0.05722755855984158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,32,0.0053342220683892565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,51200,0.04397066765361362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,65536,0.057031108273400195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,65536,0.06015466981463962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,65536,0.1373022264904446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,512,0.005964444329341252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,51200,0.045575999551349215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,64,0.0028773333049482773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,51200,0.04734222094217936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,12288,0.01571911076704661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,16384,0.018741332822375827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,16384,0.018914666440751817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,10240,0.014528888795110913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,12288,0.032455109887652926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,10240,0.01350933313369751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,128,0.005973333285914526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,10240,0.028508444627126057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,8192,0.013794666363133324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,51200,0.10940621958838569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,8192,0.011110222174061669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,12288,0.015271110667122735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,8192,0.02407466702991062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,7168,0.010803555448849997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,7168,0.02234222160445319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,6144,0.011346666349305047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,6144,0.009901333186361525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,16384,0.040692445304658674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,12288,0.014748444159825643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,5120,0.010086221827401055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,6144,0.009864889085292816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,6144,0.019606222709019978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,5120,0.00869066682126787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,5120,0.00920800036854214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,4096,0.008900444540712569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,10240,0.012791111237472959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,4096,0.00849511143234041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,4096,0.00795199970404307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,4096,0.015221332510312399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,8192,0.011760888828171624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,7168,0.012776888906955719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,3072,0.007327111230956183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,3584,0.008390222158696916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,3584,0.014125333891974555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,3072,0.0073315559162033936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,2560,0.006313777632183499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,7168,0.010701333483060202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,3584,0.008196444147162968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,2560,0.012142221960756512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,2560,0.007663999994595845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,3584,0.0074480002125104266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,2048,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,3072,0.00775822252035141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,2048,0.007296889192528195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,2560,0.006824888702895906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,1536,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,2048,0.005567111074924469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,1536,0.006948444578382704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,1024,0.004055110944641961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,1536,0.010406222608354356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,1024,0.005578666511509154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,1024,0.0064586665895250105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,1024,0.009374221993817223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,2048,0.011078221930397881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,768,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,512,0.003546666767862108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,768,0.008672888908121321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,512,0.0052764444715446895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,512,0.008310221963458592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,512,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,768,0.0038808890514903595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,16384,0.01882222294807434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,256,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,128,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,256,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,768,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,256,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,64,0.0032559999575217566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,128,0.005991111199061076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,128,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,128,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,64,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,65536,0.04376888937420315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,32,0.0052879999081293745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,65536,0.05746489100986057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,65536,0.05573778020011055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,65536,0.13721511099073622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,51200,0.03521600034501817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,32,0.002903999967707528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,51200,0.045347554816140064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,51200,0.04485866758558485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,16384,0.0151555554734336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,51200,0.10910577906502618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,16384,0.017715555098321702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,16384,0.018198221921920776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,16384,0.04076711005634732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,12288,0.015095111396577625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,12288,0.03217333224084642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,12288,0.01369599997997284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,5120,0.01773155563407474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,10240,0.012773333324326409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,8192,0.011677333050303988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,10240,0.012465777496496836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,10240,0.02815022071202596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,8192,0.01074044406414032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,8192,0.01106844428512785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,7168,0.011069333387745751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,8192,0.02446222139729394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,10240,0.012441777520709567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,7168,0.010446222292052375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,7168,0.010199111368921068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,7168,0.021374222305085924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,12288,0.01388177772363027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,6144,0.009701333112186855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,6144,0.019260444574885897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,5120,0.00868800034125646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,5120,0.017509332961506315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,5120,0.009026666482289633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,6144,0.010433778166770935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,3072,0.013134222063753339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,4096,0.008444444172912175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,6144,0.009447111023796929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,5120,0.009672889278994666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,3584,0.007640000018808577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,3584,0.01423822177780999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,3072,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,4096,0.008547555241319869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,256,0.003256888853179084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,3072,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,3072,0.007755555212497711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,2560,0.006274666637182236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,3072,0.01296266665061315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,4096,0.01515111161602868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,2048,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,2560,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,2560,0.006673777682913675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,2560,0.012134222520722283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,1536,0.004903111192915174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,2048,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,3584,0.008316444853941599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,2048,0.010785777535703448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,1024,0.004217777815130022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,1536,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,1536,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,1536,0.010080888867378235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,4096,0.007961778177155389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,3584,0.0074177781740824384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,1024,0.009357333183288574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,768,0.0058248887459437055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,768,0.006264888991912206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,512,0.0035208890007601846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,768,0.00906044410334693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,2048,0.006362666686375936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,256,0.003216888962520493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,512,0.006058666855096817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,1536,0.0068408888247278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,256,0.004973333328962326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,768,0.0038693332009845306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,128,0.003208000005947219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,256,0.006040888941950268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,256,0.007842666572994655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,128,0.004937777916590373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,128,0.008012444608741337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,32,0.0028737777223189673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,64,0.00498488876554701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,512,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,65536,0.030140442980660334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,512,0.00833777752187517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,32,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,51200,0.024639111426141527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,1024,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,65536,0.05544977717929416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,65536,0.1368888881471422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,51200,0.044569777117835156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,64,0.002891555635465516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,51200,0.10871022277408177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,16384,0.017736888594097562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,12288,0.013727111121018728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,16384,0.04063733418782552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,65536,0.055270221498277455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,1024,0.00564533347884814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,12288,0.013527111046844058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,128,0.006030222194062338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,51200,0.04348355531692505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,12288,0.031530665026770696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,10240,0.012455999851226807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,16384,0.016179554992251925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,8192,0.011403555671374003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,10240,0.02747822304566701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,10240,0.012416000167528788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,7168,0.011052444577217102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,8192,0.01108444482088089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,8192,0.010513777534166971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,12288,0.014241778188281588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,8192,0.023409777217441138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,6144,0.009355555805895064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,7168,0.010417777631017896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,16384,0.012699555191728802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,6144,0.009745777481132084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,7168,0.010128888818952773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,6144,0.0192586663696501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,5120,0.009008000294367472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,4096,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,5120,0.01738666660255856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,4096,0.00796977761718962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,10240,0.01276177747382058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,3584,0.007263999846246507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,5120,0.009015111459626092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,4096,0.015189333094490899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,5120,0.008388444781303406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,3584,0.007348444726732042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,3072,0.006599999964237213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,3584,0.014138667119873894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,2560,0.005990222096443176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,3072,0.007702222300900354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,3072,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,3072,0.01313244468635983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,2048,0.005245333330498801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,2560,0.007495111061467065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,2560,0.006688000013430913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,2560,0.012162666353914471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,1536,0.004600888739029567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,2048,0.007146666447321574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,2048,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,2048,0.011063999599880643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,1024,0.0038888889054457345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,1536,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,1536,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,1536,0.01038577738735411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,1024,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,1024,0.005611555443869696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,1024,0.009363555245929295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,7168,0.02108000053299798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,768,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,512,0.003499555504984326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,768,0.008381333616044786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,512,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,512,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,512,0.008715555899673039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,768,0.0037671112351947357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,256,0.005135111096832487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,128,0.002846222163902389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,256,0.008004444340864817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,4096,0.008360000120268928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,768,0.005437333136796951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,6144,0.009377777576446533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,128,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,32,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,128,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,768,65536,0.027273777458402846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,64,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,32,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,65536,0.053675555520587497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,65536,0.05511200096872118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,768,51200,0.02273422148492601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,65536,0.13674222098456487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,64,0.002579555536309878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,51200,0.04314577910635206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,51200,0.10859288771947224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,256,0.0031884445084465873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,3584,0.008203555312421586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,768,12288,0.011768889096048145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,16384,0.016192888220151264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,16384,0.04002666804525588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,12288,0.013824000126785703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,12288,0.013543999857372709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,12288,0.03122666809293959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,51200,0.04404088854789734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,768,16384,0.013769778112570444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,768,8192,0.010116444693671333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,10240,0.012348444097571902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,10240,0.027109333210521277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,16384,0.016878222425778706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,8192,0.010483555495738983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,8192,0.023079110516442194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,256,0.005983110931184556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,768,10240,0.010904889139864178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,768,6144,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,7168,0.010419555836253695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,7168,0.021029333273569744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,128,0.004963555683692296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,10240,0.012083555261294046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,768,5120,0.008126222425036961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,6144,0.009348444640636444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,6144,0.018924444913864136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,5120,0.008683555656009251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,768,4096,0.007814221911960179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,5120,0.009025777379671732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,7168,0.010057777994208867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,4096,0.007728888756699032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,4096,0.008436444732877944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,6144,0.009716444545321995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,3584,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,8192,0.011063111325105032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,768,3072,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,3584,0.008214222060309516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,3584,0.014118222726715935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,3072,0.007864889171388414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,768,2560,0.0059164443777667145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,768,3584,0.0074879998962084455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,3072,0.013100444442696042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,768,2048,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,2560,0.007734222544564142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,768,7168,0.009136888715955947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,2048,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,2048,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,2048,0.011408888631396823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,3072,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,1536,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,1536,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,768,1024,0.003913777983850903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,1536,0.010441777606805166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,2560,0.012479999827014076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,768,768,0.003580444388919406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,1024,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,1024,0.009593778186374241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,5120,0.017245334055688646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,768,1536,0.004612444589535396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,768,512,0.0033155555526415506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,768,0.00868977771864997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,512,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,4096,0.014848000473446317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,512,0.008519111408127679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,2560,0.006618666566080517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,1024,0.006341333190600078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,256,0.004951999833186467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,768,128,0.0028497777465316984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,256,0.008035555481910706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,768,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,128,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,768,64,0.002560888934466574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,768,32,0.0028435554769304064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,128,0.007984889050324758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,512,65536,0.02129777769247691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,512,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,64,0.005114666703674528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,32,0.005050666630268097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,512,51200,0.016132444143295288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,65536,0.05279733406172859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,256,0.005979555762476391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,51200,0.04177066683769226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,65536,0.13683999909294978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,768,0.005587555468082428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,51200,0.10845155848397149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,16384,0.016189333465364244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,16384,0.01629066632853614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,16384,0.039750221702787615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,65536,0.05454755491680569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,12288,0.01344622257683012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,12288,0.013460444079505073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,12288,0.031187554200490315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,51200,0.042486220598220825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,512,16384,0.011215111447705163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,512,8192,0.008743111458089616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,10240,0.012473777764373355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,10240,0.02753688891728719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,8192,0.01036622209681405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,512,7168,0.007735110819339752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,512,10240,0.009964444571071202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,7168,0.010034666293197209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,128,0.004948444250557157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,7168,0.010427555276287926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,10240,0.012089778151777057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,7168,0.021307556165589228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,768,256,0.003192000091075897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,6144,0.009176000124878353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,6144,0.019289778338538278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,5120,0.008357333640257517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,5120,0.01757955551147461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,5120,0.00905600024594201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,512,4096,0.007679111427730984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,512,6144,0.007329777710967594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,4096,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,4096,0.007637333538797166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,4096,0.014874666929244995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,8192,0.011149333582984077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,3584,0.00737955586777793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,512,3072,0.006604444649484422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,3584,0.008381333616044786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,3584,0.014274666706720987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,3072,0.0069546666410234236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,3072,0.007985778152942657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,512,3584,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,512,2560,0.005759999983840519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,6144,0.009735999835862054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,8192,0.023023110296991136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,512,2048,0.0052435555391841466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,2560,0.006650666809744305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,2560,0.012081777883900536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,2048,0.006275555739800136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,2048,0.007296889192528195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,512,1536,0.004576000074545543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,2048,0.010954666468832227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,1536,0.006666666517655055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,512,12288,0.009381333159075843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,512,1024,0.0038942222793896994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,1536,0.010407110883129967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,3072,0.013270222478442721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,1024,0.005607999861240387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,1024,0.009047110875447592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,768,0.006317333214812809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,512,512,0.003256888853179084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,768,0.009135999613338048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,1536,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,512,5120,0.008031999899281396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,512,0.006032889087994893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,512,256,0.0031822222388452957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,2560,0.007358222372002072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,512,0.008630221916569604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,512,768,0.003620444486538569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,512,128,0.0029306666304667792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,256,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,256,0.005192888693677055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,512,64,0.0026933333526055017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,256,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,512,32,0.002862222285734283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,768,0.005501333210203383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,256,65536,0.015950222810109455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,128,0.00508622204264005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,64,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,128,0.008052444292439355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,256,51200,0.013938667045699226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,65536,0.05178577701250712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,1024,0.006583111153708563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,51200,0.038329776790406965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,65536,0.1366480059093899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,128,0.005922666854328579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,51200,0.040896889236238264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,16384,0.01640266676743825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,32,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,65536,0.0473653342988756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,512,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,12288,0.013395555317401886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,256,10240,0.00904888908068339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,256,16384,0.009866666462686326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,51200,0.10786222087012397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,10240,0.01179377817445331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,256,8192,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,10240,0.027101332942644756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,16384,0.016561778055297006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,8192,0.01038933296998342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,256,7168,0.007837333612971837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,8192,0.02276266614596049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,12288,0.013451555536852943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,12288,0.031018667750888403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,7168,0.010393777655230628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,7168,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,7168,0.021320889393488567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,10240,0.012430222498046027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,256,5120,0.007972444097201029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,6144,0.008990222381220924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,16384,0.039044443104002215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,6144,0.019084443648656208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,256,4096,0.007293333609898885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,5120,0.008671110702885522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,5120,0.016945777667893302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,4096,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,256,3584,0.007280888656775157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,4096,0.014936000108718872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,3584,0.007305777735180325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,256,12288,0.008464888566070134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,3584,0.014121777481502958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,256,3072,0.006678222368160884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,3584,0.008008889026112026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,5120,0.009047110875447592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,256,2560,0.0063324446479479475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,3072,0.0069884442620807225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,4096,0.00835111074977451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,256,6144,0.007305777735180325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,2560,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,2560,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,2560,0.012113778127564324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,6144,0.009703999592198266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,256,1536,0.00423200014564726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,2048,0.01071999967098236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,8192,0.011067555182509951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,3072,0.007623111208279927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,1536,0.005968889014588461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,256,1024,0.003903111235962974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,1536,0.010059555371602377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,256,2048,0.005984000033802456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,1024,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,256,768,0.0035386664999855887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,1024,0.0064586665895250105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,1024,0.00903911143541336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,2048,0.006315555423498154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,768,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,768,0.005561777700980504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,768,0.009018667042255402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,1536,0.006764444626039929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,256,256,0.00295733329322603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,512,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,256,0.005183111048407025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,256,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,2048,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,256,512,0.003195555466744635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,128,0.004908444566859139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,512,0.006014222072230444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,128,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,128,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,256,32,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,64,0.004903111192915174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,128,65536,0.015441776977645027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,32,0.004959110998445087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,256,128,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,128,51200,0.013270222478442721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,65536,0.04689777890841166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,65536,0.0514284438557095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,512,0.0052782222628593445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,51200,0.0377511117193434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,51200,0.039397332403394915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,51200,0.10615999831093682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,16384,0.016201777590645682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,16384,0.016194666425387066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,256,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,12288,0.013434666726324292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,65536,0.134716444545322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,12288,0.013472889032628802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,128,12288,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,10240,0.011776000261306763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,10240,0.012418666647540199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,128,8192,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,12288,0.03094844354523553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,3072,0.01347733371787601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,8192,0.010403555300500657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,128,7168,0.0069884442620807225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,16384,0.03871555460823907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,8192,0.02307111190425025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,7168,0.010035555395815108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,7168,0.010429333481523724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,128,6144,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,10240,0.027415111660957336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,256,64,0.0028453332682450614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,6144,0.009672000176376766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,128,5120,0.00600888869828648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,6144,0.019190222024917603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,5120,0.009043555292818282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,5120,0.008402666283978356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,8192,0.011226666470368704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,5120,0.01721244388156467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,4096,0.008387555678685507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,4096,0.007701333198282454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,128,3584,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,7168,0.020959999826219346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,128,16384,0.009653333160612319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,6144,0.009175111022260454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,128,3072,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,3584,0.014176888598336114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,3072,0.007746666669845581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,128,4096,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,3072,0.01312888910373052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,2560,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,4096,0.015263999501864115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,2560,0.012104888757069906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,128,2048,0.006302222195598815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,128,10240,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,3584,0.008006222546100616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,3584,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,128,1536,0.004577777865860197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,2048,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,2048,0.006252444452709622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,128,1024,0.0038968887594011095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,3072,0.006999111009968652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,1536,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,1536,0.01020088874631458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,128,768,0.0035742223262786865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,1024,0.009354666703277165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,768,0.005241777747869492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,2560,0.0074648890230390764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,128,512,0.0032986667421129015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,768,0.009053332938088311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,2048,0.011427555647161273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,1536,0.006640888750553131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,512,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,512,0.006109333286682765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,1024,0.006334222025341458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,256,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,128,128,0.0025742221623659134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,256,0.007734222544564142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,768,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,128,2560,0.006268444574541516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,128,0.007631999750932057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,128,0.00592533333433999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,256,0.005919999960396025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,128,32,0.002846222163902389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,64,0.004937777916590373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,64,65536,0.01385155568520228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,128,0.004907555464241239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,32,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,65536,0.04894577794604831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,64,16384,0.007984889050324758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,51200,0.03956088754865859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,64,12288,0.007635555333561367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,16384,0.016181333197487723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,64,10240,0.007297777467303806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,12288,0.013343111508422427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,64,51200,0.012237333589129977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,64,7168,0.0071928890215026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,8192,0.010384000009960597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,128,256,0.002967111145456632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,64,6144,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,64,5120,0.006063999815119638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,128,64,0.0026293332791990707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,7168,0.009712888962692684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,6144,0.00902133352226681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,64,3584,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,4096,0.007648888561460707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,64,3072,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,10240,0.011904888682895236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,64,2560,0.006263111200597551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,3072,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,1024,0.005384888913896348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,2560,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,64,1536,0.004254222330119875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,5120,0.008336000144481659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,64,1024,0.0038844446341196695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,64,8192,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,1536,0.0059262220230367445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,3584,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,64,512,0.003321777822242843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,768,0.005579555614127054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,64,256,0.002862222285734283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,512,0.008438222110271454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,512,0.0052764444715446895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,256,0.005095110999213325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,64,64,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,128,0.004934222333961063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,64,32,0.002862222285734283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,64,0.00490133340160052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,1024,0.005580444302823808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,32,65536,0.013450666434235044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,32,0.004957333207130432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,64,2048,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,64,4096,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,32,16384,0.008067555725574493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,51200,0.03942666782273187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,2048,0.006311111152172089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,16384,0.01584533353646596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,64,768,0.0035964445107513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,12288,0.013456000222100152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,32,8192,0.007000889215204451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,32,7168,0.006641777853171031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,10240,0.011769777370823754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,32,51200,0.012426666915416718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,32,6144,0.006324444380071428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,32,5120,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,7168,0.010042666561073726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,64,128,0.0028817777832349143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,32,4096,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,6144,0.009173333644866943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,32,3584,0.0063902222447925145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,32,3072,0.0063982220987478895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,32,10240,0.007335999773608313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,32,2560,0.005961777849329843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,4096,0.007779555188284979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,32,2048,0.0063591111037466265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,3072,0.007076444725195567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,2560,0.006578666468461354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,32,1536,0.004621333132187526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,32,1024,0.0037626665499475268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,65536,0.04906488789452446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,2048,0.0064195555945237475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,32,768,0.0035413333939181436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,32,512,0.0032151111712058387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,5120,0.008301333420806462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,32,256,0.003055111194650332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,1024,0.0057555557125144536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,512,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,768,0.005331555588377847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,32,64,0.00254488881263468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,8192,0.010404444403118556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,32,32,0.002530666689078013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,32,12288,0.007673777639865875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,32,0.004909333255555895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,3584,0.007367110914654202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,128,32,128,0.002991111121243901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,256,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,128,0.004971555537647671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,16384,0.37209243244594997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,64,0.005123555660247803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,51200,0.6144684685601128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,1536,0.005640888793600931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,12288,0.27793865733676487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,12288,0.1507991154988607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,12288,0.10307200087441339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,51200,0.4066915512084961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,10240,0.23804177178276908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,10240,0.12713421715630427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,10240,0.08563555611504449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,51200,1.142987569173177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,16384,0.20606756210327148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,7168,0.17047466172112358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,8192,0.06957511107126872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,16384,0.13367644945780435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,7168,0.06116000148985121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,5120,0.12306222650739883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,6144,0.05360000001059639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,7168,0.09409866730372111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,8192,0.19793244202931723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,4096,0.10017510917451646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,5120,0.06663288672765096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,4096,0.03933599922392104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,4096,0.0550746652815077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,3072,0.07778044541676839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,3584,0.03472000029351976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,3584,0.04907110995716519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,2560,0.06422933604982164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,3072,0.031368000639809504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,3072,0.04416355490684509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,6144,0.07962222231758966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,5120,0.04602844516436259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,2560,0.027416000763575237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,2560,0.03735111157099406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,2048,0.023749333288934495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,2048,0.032440000110202365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,3584,0.08818399906158447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,8192,0.11058133178287083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,1024,0.029824889368481104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,1536,0.019048889478047688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,768,0.022675555613305833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,1536,0.02562933332390255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,1024,0.016664889123704698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,1024,0.019701333509551156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,768,0.016399110356966656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,256,0.010566222170988718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,2048,0.0521982212861379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,512,0.0135057775510682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,128,0.008006222546100616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,1536,0.040807111395729914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,256,0.012517333030700684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,64,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,128,0.011688888900809817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,128,0.011909333368142446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,64,0.011417778001891242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,32,0.01517688896920946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,512,0.012846222354306115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,256,0.01168444421556261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,6144,0.14526932769351536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,32,0.006061333335108227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,65536,0.6325537893507216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,65536,1.1240737703111436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,512,0.01586933268441094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,16384,0.2843617863125271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,51200,0.8784115579393176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,65536,0.401692443423801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,16384,0.10686044560538398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,12288,0.21654844284057617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,51200,0.483134216732449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,768,0.01481066644191742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,12288,0.08183377981185913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,10240,0.16776355107625326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,8192,0.13842577404446071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,12288,0.1299040052625868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,7168,0.11976799699995254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,8192,0.05665955278608534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,10240,0.1014595561557346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,7168,0.07283200158013238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,6144,0.10443555646472508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,5120,0.08926222059461807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,6144,0.04417066772778829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,51200,0.32382133271959096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,6144,0.06536710924572416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,10240,0.06889510816997953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,4096,0.07376622491412692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,3584,0.06392888890372382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,8192,0.08639822403589885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,4096,0.034833778937657676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,5120,0.05390222205056084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,16384,0.16154666741689047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,4096,0.04475288920932346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,3584,0.03979111048910353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,2560,0.047043555312686496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,3072,0.03514133228196038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,2560,0.023206222388479445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,2560,0.030517333083682593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,5120,0.03785599933730231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,2048,0.01961511042382982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,1536,0.03013422091801961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,2048,0.0269128895468182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,3072,0.0558640029695299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,3584,0.029275556405385334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,7168,0.04937066634496053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,3072,0.026276444395383198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,768,0.017135999268955655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,1024,0.01367644468943278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,1024,0.016294222739007738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,2048,0.03978311022122701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,768,0.012866666747464074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,768,0.013909333282046847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,256,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,512,0.011403555671374003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,512,0.011905777785513135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,128,0.00655377780397733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,1536,0.016210666961140104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,1536,0.021361778179804485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,64,0.005642666584915585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,128,0.010495111346244812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,32,0.006269333263238271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,64,0.011765333513418833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,32,0.013010666602187686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,65536,0.32301423284742564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,256,0.010407110883129967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,1024,0.024018666810459558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,51200,0.2504462136162652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,65536,0.14255466726091173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,512,0.01238844460911221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,16384,0.08772177828682794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,51200,0.17512888378567168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,12288,0.06524266799290974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,51200,0.11087910996543036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,16384,0.0415653321478102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,10240,0.055890666113959424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,12288,0.032953778902689614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,8192,0.047341333495246045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,12288,0.04622488882806566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,10240,0.03966666592491998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,7168,0.04008977942996555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,8192,0.023992000354660883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,8192,0.03457955519358317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,6144,0.03560355636808608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,7168,0.021540444758203294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,7168,0.02987733483314514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,5120,0.031047112411922876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,6144,0.019335110982259113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,16384,0.06144622299406263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,6144,0.026499556170569524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,4096,0.026622222529517278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,5120,0.018771555688646104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,3584,0.02308533257908291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,256,0.010604444477293225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,4096,0.01516888870133294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,4096,0.019352888067563374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,128,0.010092444717884064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,3584,0.014512000812424554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,2560,0.017508443858888414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,3072,0.013063111239009433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,3072,0.01574577722284529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,2048,0.01424088908566369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,2560,0.012091555529170565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,2560,0.013822221921549903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,1536,0.011609777808189392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,2048,0.011108444796668159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,2048,0.012439111040698158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,5120,0.023042667243215773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,65536,0.2305386596255832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,1536,0.01070844464831882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,3072,0.02020711037847731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,768,0.00795022232664956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,1024,0.008664888640244802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,1024,0.00927111092540953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,768,0.007648000286685095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,768,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,256,0.004508444418509801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,512,0.007349333001507654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,512,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,128,0.003997333347797394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,256,0.0069520001610120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,256,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,1024,0.009226666556464301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,10240,0.027862221002578735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,64,0.003882666842805015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,128,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,3584,0.017645332548353408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,32,0.007836444510353936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,65536,0.2406568792131212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,65536,0.17545599407619902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,65536,0.12524978319803873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,1536,0.00942133367061615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,16384,0.0658515559302436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,51200,0.1347528828514947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,512,0.006015111174848344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,12288,0.05084088775846693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,64,0.006635555376609166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,16384,0.048617776897218495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,10240,0.04243377844492594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,12288,0.02987733483314514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,51200,0.18855288293626574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,8192,0.035459554857677884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,10240,0.031896889209747314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,128,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,51200,0.10001777940326267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,7168,0.031074666314654883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,16384,0.03760622276200189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,7168,0.0199626667631997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,6144,0.027242667145199243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,10240,0.025674665967623394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,6144,0.018055111169815063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,32,0.0039057777159743835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,5120,0.023696889479955036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,8192,0.027632888820436265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,4096,0.020676443974177044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,8192,0.022308443983395893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,5120,0.01827555563714769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,4096,0.014471999473041959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,4096,0.01572177807490031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,3584,0.013788444300492605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,3584,0.014404444230927361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,6144,0.021200888686709937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,3072,0.013226666384273104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,2560,0.013456888496875763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,3584,0.017808889349301655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,2560,0.01161688897344801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,12288,0.03683111071586609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,2560,0.011852444046073489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,1536,0.009956444303194681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,2048,0.010030222435792288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,3072,0.01571466690964169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,2048,0.010958222051461538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,1536,0.008799999952316284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,1536,0.009630222287442949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,768,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,1024,0.008001777860853408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,1024,0.007691555553012424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,5120,0.017087111870447796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,512,0.004899555610285865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,768,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,256,0.0042160000238153665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,512,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,2048,0.011752888560295105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,512,0.006322666588756774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,256,0.006856888946559694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,256,0.006054222169849608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,64,0.003543111185232798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,1024,0.008067555725574493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,128,0.006738666858938005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,3072,0.012839111189047495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,64,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,32,0.007149332927332983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,65536,0.20390933089786104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,768,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,7168,0.024315555890401203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,51200,0.15693777137332493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,128,0.0038924444880750445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,65536,0.12070222695668538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,51200,0.0969688892364502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,32,0.003875555677546395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,16384,0.04617066515816582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,12288,0.02800977892345852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,12288,0.0350524451997545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,65536,0.16773245069715711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,10240,0.03047200043996175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,8192,0.031067556805080835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,10240,0.024759110477235582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,51200,0.1296444469028049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,128,0.005987555616431766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,8192,0.021106666988796655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,8192,0.02647200061215295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,6144,0.0229768885506524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,7168,0.023038221730126276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,6144,0.01757244434621599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,6144,0.02014311154683431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,5120,0.020092444287406076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,16384,0.05368266503016154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,5120,0.0162426663769616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,4096,0.01719199948840671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,5120,0.017528888252046373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,3584,0.014735110931926303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,4096,0.014512888259357877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,4096,0.01520088811715444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,3072,0.013329777452680798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,3584,0.014064888159434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,3072,0.012288000020715924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,3072,0.013095999757448832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,16384,0.03609688745604621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,2560,0.011731555892361535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,2048,0.010015111002657149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,2560,0.01075644459989336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,7168,0.020094222492641874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,1536,0.008645333349704742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,2048,0.009395555489593083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,2048,0.010710222025712332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,1024,0.007338666253619724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,1536,0.008475555313958062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,1536,0.008761777646011775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,768,0.005590222362014983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,1024,0.007831999825106727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,1024,0.0073724447025193115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,7168,0.026133333643277485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,768,0.007357333269384172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,768,0.00666133314371109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,10240,0.035622222556008235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,256,0.003846222327815162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,2560,0.011427555647161273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,256,0.005976000179847081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,128,0.0035866668654812705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,512,0.004504888835880491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,128,0.006803555621041193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,64,0.0032275555034478507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,3584,0.01313066648112403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,32,0.0035626664757728577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,512,0.006313777632183499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,64,0.006058666855096817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,65536,0.17892889181772867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,32,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,256,0.006665777828958299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,51200,0.14116177293989393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,65536,0.12474666701422797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,16384,0.04852799905671013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,128,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,51200,0.05805866585837471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,51200,0.09648355510499741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,12288,0.038190222448772855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,16384,0.023372444841596816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,10240,0.03188088867399428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,512,0.007019555403126611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,12288,0.018840889135996502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,8192,0.02725244396262699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,10240,0.016567111015319824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,7168,0.023987554841571387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,8192,0.014757333530320061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,8192,0.0207022229830424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,6144,0.020958221620983545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,12288,0.043471111191643604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,65536,0.07173244158426921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,5120,0.01827022267712487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,6144,0.012720888687504662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,6144,0.016185777054892648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,4096,0.01549600064754486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,16384,0.035395556026034884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,5120,0.01420088940196567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,3584,0.01275022245115704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,4096,0.010852444503042432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,4096,0.012768000364303589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,3072,0.011720889144473605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,3584,0.009679999616410997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,3584,0.011748444702890186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,7168,0.01368799971209632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,2560,0.01036888857682546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,3072,0.00866844422287411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,7168,0.017665778597195942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,2048,0.009208888643317753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,2560,0.008364444805516137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,1536,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,2560,0.010021333065297868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,5120,0.011591999895042844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,1024,0.006137777947717243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,1536,0.007615110940403408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,12288,0.02736088964674208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,10240,0.023873777853118047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,1024,0.006640000061856375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,1024,0.006738666858938005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,512,0.004583111239804162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,768,0.00664444433318244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,768,0.006275555739800136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,3072,0.010950222611427307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,256,0.003850666599141227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,512,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,128,0.003547555456558863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,256,0.006112888869312074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,2048,0.009071110851234859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,256,0.005587555468082428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,64,0.003551111039188173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,128,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,32,0.0035537779331207275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,128,0.005564444594913059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,768,0.004954666727119022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,65536,0.16230044100019667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,64,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,32,0.005970666805903117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,51200,0.12888177235921225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,65536,0.12244089444478352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,16384,0.05041422115431892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,51200,0.05611999829610189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,512,0.005782222168313132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,16384,0.03446755475468106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,1536,0.007294221884674496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,10240,0.029771556456883747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,10240,0.02271644439962175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,65536,0.06902666886647542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,10240,0.016186666157510545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,51200,0.09415288766225179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,2048,0.007684444387753804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,7168,0.022340445054901972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,8192,0.019645333290100098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,12288,0.01792355544037289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,6144,0.019567999574873183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,7168,0.017258667283587985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,6144,0.012459555433856117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,5120,0.016767111089494493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,8192,0.025419554776615564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,12288,0.03540444374084473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,8192,0.014336888988812765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,4096,0.014884443746672736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,5120,0.014150222142537435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,12288,0.02647999922434489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,3584,0.012141333685980903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,4096,0.00997511131895913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,4096,0.012704888979593912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,3072,0.011034666664070554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,3584,0.00923644420173433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,3584,0.011671110987663269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,2560,0.010036444498433007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,3072,0.008485332959228093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,3072,0.010760000182522668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,16384,0.02203911046187083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,2048,0.00887288898229599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,2560,0.009705777797434065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,5120,0.011408000356621213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,1536,0.0074471111098925276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,2048,0.00868622213602066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,7168,0.01330044451687071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,1536,0.007024889190991719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,1536,0.007376000285148621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,768,0.005094222310516569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,1024,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,1024,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,512,0.004231111043029361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,768,0.006002666635645761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,2560,0.00794933322403166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,512,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,512,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,2048,0.007340444458855524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,128,0.003589333345492681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,256,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,1024,0.005568888866239124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,64,0.0035342222286595237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,128,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,128,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,6144,0.015954666667514376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,65536,0.1436391141679552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,32,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,51200,0.11354222562577988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,256,0.003920888735188378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,51200,0.08981155686908299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,51200,0.054105778535207115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,16384,0.04494488901562161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,16384,0.021597334080272253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,768,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,65536,0.1157733334435357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,12288,0.025387555360794067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,10240,0.024142222272025213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,12288,0.01754488878779941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,256,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,32,0.003547555456558863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,10240,0.01570755574438307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,10240,0.021704889006084863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,7168,0.018177777528762817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,64,0.005636444522274866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,12288,0.029361777835422095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,8192,0.019389332996474374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,6144,0.015855999456511605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,7168,0.013067555924256643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,7168,0.01696622206105126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,6144,0.012061333490742577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,6144,0.015451555450757345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,4096,0.012409778104888068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,5120,0.010767111347781287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,5120,0.013759111364682516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,3584,0.01105955574247572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,4096,0.009572444690598382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,4096,0.012302222351233164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,3072,0.00999466660949919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,8192,0.014076444837782117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,16384,0.033018665181265935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,2560,0.008873778084913889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,3072,0.00835111074977451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,5120,0.014069333672523499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,2048,0.007661333514584436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,2560,0.00776800016562144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,2560,0.009138666921191746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,2048,0.007408888803588019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,2048,0.007991111112965478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,8192,0.020894222789340552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,1024,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,1536,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,1536,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,768,0.004573333180612988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,1024,0.006613333192136552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,1024,0.006641777853171031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,512,0.004222222086456087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,3072,0.010696888797812991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,768,0.006241777704821692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,256,0.0035413333939181436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,512,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,1536,0.0068488890926043195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,128,0.0032826666202810076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,256,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,3584,0.008993777963850234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,3584,0.011415999796655444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,128,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,768,0.006518222391605377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,65536,0.06648266977734037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,32,0.0032239999208185407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,65536,0.11768533123864068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,512,0.005623111294375525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,256,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,32,0.005984888722499211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,51200,0.09540444612503052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,65536,0.06407555606630114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,128,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,16384,0.0388008885913425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,51200,0.052036444346110024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,12288,0.03091199861632453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,16384,0.031472888257768415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,64,0.005672889037264719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,12288,0.016861332787407767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,12288,0.024175110790464614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,10240,0.015233778291278414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,10240,0.021012443635198805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,8192,0.01924977699915568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,65536,0.11065600315729777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,64,0.0031964443624019623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,7168,0.016728000508414376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,8192,0.013417777915795645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,8192,0.018560000591807894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,7168,0.012472888661755456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,5120,0.01292711082432005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,6144,0.01149511088927587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,6144,0.015093333191341825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,10240,0.02753422160943349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,4096,0.011495999991893768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,5120,0.010030222435792288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,3584,0.010099555883142684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,4096,0.009282666775915358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,4096,0.011977777712874942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,3584,0.008385777473449707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,3072,0.009196444518036312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,3584,0.01094044413831499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,51200,0.08629600207010905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,3072,0.008079999850855933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,16384,0.02030666669209798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,3072,0.009855111440022787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,7168,0.016554666890038382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,2560,0.007772444850868649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,2560,0.00869333330127928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,2048,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,2048,0.007696888513035244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,5120,0.013451555536852943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,1536,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,1536,0.0070159998204973005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,768,0.004561777744028303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,1024,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,1024,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,512,0.0038977778620190094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,768,0.006320888797442119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,768,0.006156444549560547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,2048,0.007640000018808577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,256,0.003524444583389494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,512,0.006161777923504512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,128,0.003548444559176763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,1536,0.006301333506902059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,256,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,64,0.0031928889867332247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,128,0.006039999839332368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,32,0.003371555565132035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,128,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,65536,0.09214222431182861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,6144,0.014555555250909595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,32,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,65536,0.06957955492867364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,512,0.005707555760939916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,256,0.005325333525737126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,1024,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,51200,0.05433244506518046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,64,0.005640888793600931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,16384,0.02126755482620663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,12288,0.025528889563348558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,2560,0.00831733312871721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,10240,0.02239377796649933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,12288,0.016523554921150208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,12288,0.016524444023768108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,65536,0.062068442503611244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,8192,0.016493333710564505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,10240,0.014714666538768344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,7168,0.014146667387750415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,8192,0.013090666797426013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,16384,0.03200711144341363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,8192,0.013556444810496436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,7168,0.011863999896579318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,16384,0.019869332512219746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,5120,0.011420444481902652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,6144,0.011047999891969891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,5120,0.009352000223265754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,4096,0.010694222317801582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,51200,0.07641600237952338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,4096,0.0087013335691558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,4096,0.00903911143541336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,3584,0.009043555292818282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,3584,0.008343111309740279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,6144,0.01071200023094813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,3072,0.008647999829716152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,5120,0.010055111514197456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,10240,0.014742222097184924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,3072,0.007667555577225155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,2560,0.007484444313579136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,3072,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,2048,0.006768888897365994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,2560,0.007704888780911763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,6144,0.012693333129088083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,2560,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,2048,0.007335999773608313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,7168,0.012075555821259817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,2048,0.0064444442590077715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,1024,0.0042444442709287005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,1536,0.006999111009968652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,1536,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,1024,0.006544888847404056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,1024,0.005659555395444234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,768,0.006268444574541516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,768,0.005620444400442972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,51200,0.04953333404329088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,256,0.003237333355678452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,1536,0.005587555468082428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,256,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,128,0.0033653332955307434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,768,0.003908444609906939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,128,0.005998222364319696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,512,0.003654222107595868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,64,0.003126222226354811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,32,0.0032319999817344877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,65536,0.08498044146431817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,512,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,32,0.005607999861240387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,51200,0.0696622199482388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,65536,0.06088711155785454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,65536,0.06673955255084567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,256,0.005221333354711533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,51200,0.04877422253290812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,51200,0.05244000090493096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,128,0.004967111266321606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,16384,0.020047111643685233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,512,0.005983110931184556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,12288,0.016202666693263583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,10240,0.02168088820245531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,12288,0.01605866683853997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,8192,0.015507555670208402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,10240,0.01457155578666263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,3584,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,10240,0.014190221826235453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,16384,0.02852533260981242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,8192,0.012987555729018318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,6144,0.012135999898115793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,7168,0.011726222104496427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,12288,0.024392889605628118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,5120,0.010910222099887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,6144,0.01072266697883606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,64,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,4096,0.010011555420027839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,5120,0.009376889301670922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,5120,0.009692444569534725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,3584,0.008831111093362173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,7168,0.013786666923099093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,8192,0.012763555679056378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,3072,0.00831733312871721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,3584,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,7168,0.011711111499203576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,3584,0.007916444705592262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,3072,0.008008889026112026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,3072,0.007287110719415877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,2048,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,2560,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,2048,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,4096,0.008709333009190029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,4096,0.008622222476535374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,1536,0.0069511110583941145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,2560,0.007318221860461765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,1024,0.0042257776690853965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,16384,0.0195795562532213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,768,0.0038853333228164246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,6144,0.01036711119943195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,1024,0.006602666858169768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,512,0.0035422220826148987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,1024,0.00564533347884814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,768,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,256,0.003536000019974179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,512,0.005981333139869902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,2048,0.007196444604131911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,1536,0.005142222262091107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,128,0.003265777809752358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,512,0.00535999983549118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,256,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,64,0.0029493332323100832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,128,0.005997333261701796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,128,0.004931555440028508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,65536,0.07632178068161011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,2560,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,32,0.0053102220926019884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,768,0.005629333357016246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,65536,0.05961600277158949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,51200,0.06195288896560669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,256,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,51200,0.04761066701677111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,32,0.003212444484233856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,1536,0.006157333238257303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,12288,0.021254221598307293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,64,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,16384,0.019545776976479422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,10240,0.018609777092933655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,12288,0.01607466737429301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,12288,0.015631111131774057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,65536,0.06437244680192736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,10240,0.014087110757827759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,10240,0.013834666874673633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,8192,0.012180444267061023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,51200,0.05059822069274055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,7168,0.01073688848151101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,16384,0.019292443990707397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,6144,0.010602666272057427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,8192,0.01657777859105004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,5120,0.010577778021494547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,5120,0.009341333475377824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,7168,0.014446222119861178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,4096,0.009513777991135916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,4096,0.008672888908121321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,3584,0.008536888493431939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,4096,0.008299555215570662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,3584,0.008366222182909647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,3584,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,6144,0.010047111246320937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,3072,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,2560,0.007287110719415877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,16384,0.025332444243960913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,2560,0.007692444655630324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,2048,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,2048,0.006330666856633292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,2048,0.007340444458855524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,1536,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,3072,0.007303111255168915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,8192,0.012474666866991254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,1024,0.004579555657174853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,1536,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,1536,0.006045333213276333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,7168,0.011415111521879831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,1024,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,1024,0.005595555735958948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,512,0.0038737778862317405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,3072,0.007855110698276097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,512,0.006032889087994893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,256,0.0032524443748924467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,5120,0.009363555245929295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,256,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,128,0.0033777778347333274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,2560,0.006691555596060223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,768,0.0042053332759274375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,64,0.0029022221763928733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,128,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,32,0.0030257778449190986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,768,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,65536,0.0635591083102756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,512,0.005247999810510212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,32,0.005244444641802046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,51200,0.05348799957169426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,65536,0.05864355299207899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,51200,0.04863022102249993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,16384,0.02257066633966234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,51200,0.04639288783073425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,768,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,16384,0.01902222302224901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,64,0.0053022222386466126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,12288,0.01533422205183241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,12288,0.01588355501492818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,10240,0.017047999633683097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,65536,0.061497781011793345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,8192,0.01516177753607432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,10240,0.013658666776286231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,8192,0.011432888607184092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,12288,0.01923466722170512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,7168,0.011052444577217102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,7168,0.010784889260927836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,6144,0.01276088919904497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,8192,0.012072000238630505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,6144,0.009932444327407414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,5120,0.01037599974208408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,10240,0.013768889009952545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,256,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,4096,0.00905511114332411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,5120,0.009108444054921469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,16384,0.018932445181740653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,3584,0.008314666648705801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,3584,0.007636444436179266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,5120,0.009022222624884712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,3584,0.00833422193924586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,3072,0.007710222568776872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,3072,0.00794400026400884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,2560,0.006565333240562015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,4096,0.008202667037645975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,2560,0.007756444315115611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,2048,0.005969777703285217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,6144,0.011767110890812345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,2048,0.006337777607970768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,2048,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,1536,0.0052426668504873914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,2560,0.00674933319290479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,1536,0.006057777752478917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,1536,0.006952889263629913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,1024,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,128,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,1024,0.0063368889192740125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,768,0.005628444254398346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,1024,0.004282666577233208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,768,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,512,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,6144,0.010377777947319878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,768,0.0042453333735466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,512,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,128,0.002924444360865487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,256,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,128,0.0059057776298787855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,128,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,64,0.0028657778683635923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,32,0.0032222221295038858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,4096,0.00867022242810991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,256,0.0032222221295038858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,32,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,256,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,65536,0.0572320024172465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,51200,0.04378222094641792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,512,0.003916444463862313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,51200,0.04524711105558607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,16384,0.01903288894229465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,51200,0.04668177829848396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,12288,0.01551644504070282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,16384,0.018731556004948087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,7168,0.012856889102194043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,10240,0.013790222505728403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,12288,0.014298665854665967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,12288,0.014912888407707214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,8192,0.013576000101036496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,10240,0.01311288856797748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,65536,0.05888711081610786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,7168,0.012079111403889127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,8192,0.011335999601417117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,7168,0.010616000327799054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,7168,0.01071200023094813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,3072,0.007274666594134436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,65536,0.05442666676309374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,5120,0.010041777458455827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,6144,0.009694221946928237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,5120,0.009108444054921469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,4096,0.008647999829716152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,5120,0.00868266655339135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,8192,0.011576889289749993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,3584,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,4096,0.008676444490750631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,4096,0.007842666572994655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,3584,0.008279110822412703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,3584,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,3072,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,2560,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,3072,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,2560,0.007640889121426477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,2048,0.005579555614127054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,2560,0.006609777609507243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,6144,0.011528888510333167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,2048,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,64,0.005280888742870755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,1536,0.006828444285525217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,1024,0.00423555572827657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,10240,0.013501333693663279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,1024,0.006545777950021956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,768,0.003759110967318217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,3072,0.007254222200976477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,2048,0.0063271112740039825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,512,0.0035208890007601846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,6144,0.009702222214804756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,512,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,512,0.005286222116814719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,256,0.003185777821474605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,1024,0.005655111124118169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,256,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,256,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,1536,0.004616888860861461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,128,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,128,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,32,0.0028560000161329904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,1536,0.006021333237489064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,64,0.005072888814740711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,32,0.005218666460778978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,65536,0.05578133132722643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,65536,0.05628444751103719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,51200,0.034761776526769005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,64,0.002867555452717675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,16384,0.01833599971400367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,16384,0.015449777245521545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,16384,0.017940445078743827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,65536,0.04272800021701389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,16384,0.017893332574102614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,12288,0.013803555733627744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,10240,0.012496888637542725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,51200,0.044734223021401294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,768,0.005590222362014983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,51200,0.04479733440611097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,8192,0.011411555939250521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,10240,0.012470222181744046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,10240,0.012821333275900947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,7168,0.011320888996124268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,8192,0.01071111112833023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,128,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,7168,0.010560000108347999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,7168,0.010048888623714447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,6144,0.009709333380063375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,5120,0.009651555783218807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,768,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,5120,0.009009777671760982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,4096,0.008665777742862701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,5120,0.00849422232972251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,8192,0.011076444553004371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,3584,0.007274666594134436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,4096,0.00831911133395301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,12288,0.014490667316648694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,4096,0.007972444097201029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,6144,0.010425777898894416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,3584,0.008012444608741337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,2560,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,3072,0.00796622203456031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,3072,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,2048,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,2560,0.00758133331934611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,2560,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,1536,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,2048,0.0070364442136552595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,1024,0.004219555606444676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,1536,0.006868444383144379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,1536,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,768,0.00387999994887246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,1024,0.006532444308201472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,1024,0.005639111002286275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,3584,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,768,0.006274666637182236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,768,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,256,0.0032186667538351486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,512,0.005311111195219888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,12288,0.013565333353148567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,2048,0.0062977779242727495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,3072,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,512,0.003531555748648114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,128,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,6144,0.009340444372759925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,512,0.005990222096443176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,64,0.004940444396601783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,256,0.005981333139869902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,256,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,65536,0.05508977837032742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,51200,0.02591466738118066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,128,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,64,0.0028968888024489083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,16384,0.012881778180599213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,51200,0.044328000810411244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,16384,0.016183111402723525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,16384,0.017539555827776592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,128,0.0031724443866146933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,12288,0.013808888693650564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,12288,0.013470222552617392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,65536,0.053865777121649853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,10240,0.012014221813943652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,51200,0.04275555411974589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,8192,0.011215111447705163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,8192,0.011062222222487131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,32,0.005091555416584015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,10240,0.01255644443962309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,7168,0.010042666561073726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,6144,0.009580444130632613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,7168,0.010435555544164447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,10240,0.012422222230169507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,6144,0.0097884444726838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,5120,0.009012444151772393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,6144,0.00923466682434082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,4096,0.007975999679830339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,5120,0.00904088881280687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,7168,0.01072622256146537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,5120,0.008643555144468943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,4096,0.008364444805516137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,4096,0.007841777470376756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,3072,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,3584,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,3584,0.00742222203148736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,2560,0.006026666611433029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,3072,0.007679111427730984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,3072,0.006979555719428592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,32,0.0032097777972618737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,2048,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,2560,0.006670222100284364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,2048,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,12288,0.013783999615245394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,2048,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,3584,0.00741244438621733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,1536,0.006792888873153263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,1536,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,768,0.0038942222793896994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,1024,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,2560,0.007518221934636434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,768,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,512,0.003552000141806073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,768,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,65536,0.03199377655982971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,8192,0.01071733319097095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,1024,0.004238222208287981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,512,0.005967999911970562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,512,0.0052933332820733385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,256,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,1024,0.0063582224150498705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,64,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,128,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,128,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,64,0.005199999858935674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,768,65536,0.026644445127911035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,1536,0.00462311092350218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,256,0.0032293332947625052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,768,51200,0.02223555578125848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,65536,0.054493334558275014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,65536,0.05298044284184774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,768,16384,0.013502221968438892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,51200,0.04378755556212532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,768,12288,0.012069332930776807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,51200,0.04228444563017952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,32,0.0026702222724755607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,768,10240,0.01106844428512785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,12288,0.01350044459104538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,12288,0.013447999954223633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,768,8192,0.010056888891590966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,10240,0.011902222202883827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,8192,0.011060444845093621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,128,0.0028942223224374983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,768,7168,0.009520000053776635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,256,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,768,6144,0.008671999805503422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,16384,0.01626844373014238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,7168,0.009834667046864828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,768,5120,0.007848888635635376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,6144,0.009716444545321995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,6144,0.009112888740168678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,768,4096,0.00797155582242542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,10240,0.0123831108212471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,5120,0.008661333057615492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,768,3584,0.007400888535711501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,4096,0.008394666843944127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,4096,0.0076657773719893555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,768,3072,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,3584,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,3584,0.0073520003093613525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,3072,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,16384,0.01686311099264357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,2560,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,2560,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,32,0.005115555392371284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,2048,0.0070017774899800616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,2048,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,8192,0.010378667049937779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,1536,0.006654222392373615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,768,1024,0.003903111235962974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,3072,0.0069884442620807225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,1024,0.006274666637182236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,768,2048,0.005208000126812193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,5120,0.009087111386987898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,768,0.006275555739800136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,768,1536,0.004616888860861461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,7168,0.01038755559258991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,512,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,768,256,0.0032097777972618737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,768,768,0.003879111260175705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,256,0.004981333182917701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,768,128,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,768,512,0.0032711111836963226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,128,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,768,64,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,768,2560,0.005912000106440649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,768,32,0.0029022221763928733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,64,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,1024,0.005622222191757626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,512,65536,0.020000000794728596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,32,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,512,51200,0.01684888866212633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,65536,0.05394755469428169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,768,0.005567111074924469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,65536,0.051623112625545926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,512,16384,0.011048888994587792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,51200,0.039112889104419284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,128,0.005970666805903117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,16384,0.016188444362746347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,1536,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,16384,0.016057777735922072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,512,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,12288,0.013481777575280933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,10240,0.012415111064910889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,10240,0.011745777395036487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,256,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,512,7168,0.008043555749787224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,8192,0.010421333213647207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,7168,0.010415111151006486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,7168,0.009700444837411245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,512,6144,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,512,10240,0.009733333355850643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,512,8192,0.008908444808589088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,6144,0.009006222089131674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,512,5120,0.00870311094654931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,5120,0.00905511114332411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,512,4096,0.007688889073001013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,5120,0.00835111074977451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,51200,0.04178222351604038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,512,3584,0.006994666324721442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,4096,0.007697777615653143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,12288,0.013627555635240344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,512,3072,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,3584,0.008032889001899296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,6144,0.009732444253232744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,512,2560,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,3072,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,3072,0.0069751110341813825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,8192,0.011008000208271874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,2560,0.007375111182530721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,2560,0.00665511108107037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,512,1536,0.004580444345871608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,2048,0.007322666545708974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,512,12288,0.009733333355850643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,2048,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,1536,0.006710222197903528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,1536,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,512,768,0.0035457776652442086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,1024,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,3584,0.007335999773608313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,512,512,0.0032195556494924757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,768,0.006063111126422882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,512,256,0.0031991110493739447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,512,2048,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,512,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,512,128,0.0028968888024489083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,256,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,256,0.005099555684460534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,512,64,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,128,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,512,32,0.0028702221396896574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,128,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,256,65536,0.015834665960735746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,1024,0.006592888798978593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,64,0.00490044429898262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,256,51200,0.014126222994592456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,65536,0.047545777426825635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,65536,0.05099111133151584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,512,0.006014222072230444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,51200,0.03801688883039687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,51200,0.03906666570239597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,256,12288,0.008332444561852349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,16384,0.016193777322769165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,512,1024,0.003927111211750242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,256,10240,0.01109155515829722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,12288,0.01348977784315745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,12288,0.01312888910373052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,256,8192,0.009714666340086196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,10240,0.01239999963177575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,10240,0.011752888560295105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,256,7168,0.009472889204819998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,8192,0.01110755569405026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,768,0.005573333137565189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,256,16384,0.009690666364298927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,7168,0.010474666953086853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,7168,0.009714666340086196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,256,5120,0.008714666797055138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,16384,0.016177778442700703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,5120,0.00906399968597624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,5120,0.008358221915033128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,32,0.004942222187916438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,8192,0.010379555324713389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,4096,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,4096,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,3584,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,256,3072,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,3584,0.007985778152942657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,256,4096,0.007369777394665613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,256,2560,0.0059057776298787855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,3072,0.0069546666410234236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,2560,0.007361777954631382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,256,6144,0.008671999805503422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,2560,0.006642666541867786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,2048,0.007156444092591603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,6144,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,3072,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,1536,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,256,1024,0.00387999994887246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,1536,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,4096,0.008383110993438298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,256,3584,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,1024,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,6144,0.009768000079525841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,768,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,256,1536,0.004531555705600315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,512,0.00600355573826366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,256,256,0.0028853333658642242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,512,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,256,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,256,0.005129777722888523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,256,128,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,2048,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,128,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,256,64,0.002522666628162066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,64,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,256,32,0.0028159999185138275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,256,512,0.0032968889507982465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,128,65536,0.015796444482273526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,65536,0.04658755660057068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,65536,0.04862222075462341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,256,2048,0.005010666532648934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,51200,0.03872177667087979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,51200,0.037695109844207764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,128,16384,0.009755555954244401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,768,0.005285333428117964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,128,12288,0.00980533328321245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,16384,0.01622488929165734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,16384,0.01603200038274129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,128,10240,0.0090524446633127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,12288,0.013518222504191928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,128,51200,0.015203555425008139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,128,8192,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,10240,0.012443555725945367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,256,768,0.0035742223262786865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,128,7168,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,8192,0.011064888702498542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,8192,0.010416000253624385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,128,0.004963555683692296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,7168,0.009723555710580613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,6144,0.009699555734793345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,6144,0.009081777599122789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,12288,0.01310488912794325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,5120,0.00903200027015474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,1024,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,128,4096,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,4096,0.008347555167145198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,128,3584,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,3584,0.007272889216740926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,128,5120,0.007682667010360294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,128,3072,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,3584,0.007840000092983246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,10240,0.011747555600272285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,5120,0.008377778033415476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,128,2560,0.005653333332803514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,3072,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,128,2048,0.00490399988161193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,2560,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,32,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,2560,0.006678222368160884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,2048,0.007009777757856581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,2048,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,128,1024,0.0038808890514903595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,1536,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,1536,0.005983110931184556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,128,768,0.0035582222044467926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,1024,0.006320888797442119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,1024,0.005580444302823808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,128,512,0.0032008888406885993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,768,0.006270222365856171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,768,0.005623999983072281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,128,256,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,7168,0.010422222316265106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,128,1536,0.004576000074545543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,128,128,0.0028364445186323593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,256,0.005922666854328579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,256,0.005233777893914117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,128,64,0.0028328889360030494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,128,0.00599644457300504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,128,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,64,0.005008000052637524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,3072,0.007666666474607255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,32,0.004909333255555895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,64,51200,0.014031110538376702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,512,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,64,16384,0.009314667019579146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,65536,0.047932446002960205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,64,12288,0.009200000100665623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,51200,0.038725333081351385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,64,10240,0.00868800034125646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,16384,0.015987555185953777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,64,8192,0.007975111405054728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,128,32,0.002899555489420891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,12288,0.01313333296113544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,64,7168,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,10240,0.011742221812407175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,64,65536,0.014854222536087036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,8192,0.010372444159454769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,64,5120,0.007696888513035244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,7168,0.009475555684831407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,6144,0.00902844468752543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,64,3584,0.006789333290523953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,5120,0.00833422193924586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,64,3072,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,4096,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,64,2560,0.006308444258239534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,3584,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,64,2048,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,3072,0.006955555743641323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,64,1536,0.004546666724814309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,2560,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,2048,0.006302222195598815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,64,4096,0.006990222467316522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,128,6144,0.008656000097592672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,64,768,0.003555555724435382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,1024,0.005647111270162795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,64,512,0.003175111073586676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,768,0.0053075556125905775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,64,256,0.0028524444335036804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,64,6144,0.008375999828179678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,512,0.005292444593376583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,64,128,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,64,64,0.002598222138153182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,256,0.005114666703674528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,64,32,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,64,1024,0.0038951109680864546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,4096,0.007660444411966536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,64,0.004960888789759742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,32,51200,0.021028444170951843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,65536,0.04793599910206265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,32,16384,0.009352888498041365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,51200,0.03886311252911886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,32,12288,0.009704888694816166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,16384,0.015856888559129503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,32,10240,0.00867377801073922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,512,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,32,8192,0.008056888977686564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,32,7168,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,10240,0.011576889289749993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,32,6144,0.008277333445019191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,8192,0.010418666733635796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,32,5120,0.007838221887747446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,6144,0.009027555584907532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,32,4096,0.0069751110341813825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,128,0.004980444494220945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,12288,0.013172444370057849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,32,3584,0.007003555695215861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,32,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,7168,0.009692444569534725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,32,2560,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,3072,0.0069884442620807225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,32,2048,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,2560,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,1536,0.005657777604129579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,4096,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,2048,0.006311111152172089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,32,3072,0.006648889018429651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,32,768,0.0035706667436493766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,1536,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,1024,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,768,0.005321777943107817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,5120,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,32,1536,0.004237333519591225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,32,128,0.0025395556456512874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,3584,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,128,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,32,64,0.002552888873550627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,32,32,0.002543111021320025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,64,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,32,0.00490399988161193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,32,65536,0.025440000825458105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,256,0.005245333330498801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,32,512,0.003248888792263137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,51200,0.612135092417399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,51200,1.1417475806342232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,16384,0.3745768864949544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,32,1024,0.00388711111413108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,16384,0.20326222313774955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,51200,0.39918311436971027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,16384,0.13600711027781168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,512,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,12288,0.1067919996049669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,10240,0.24234933323330352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,10240,0.1288426717122396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,12288,0.2830293443467882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,8192,0.19295556015438506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,10240,0.08587466345893012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,12288,0.2216471168729994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,8192,0.06928977701399061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,8192,0.10748889048894246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,7168,0.17053422662946913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,7168,0.09256711271074082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,8192,0.15523733033074275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,113,32,256,0.002883555574549569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,6144,0.14516711235046387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,7168,0.06120622158050537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,10240,0.19387466377682158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,6144,0.08210400078031752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,5120,0.12485067049662273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,5120,0.06684444348017375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,6144,0.11813955836825901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,12288,0.1500764422946506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,4096,0.09865955511728923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,5120,0.045928888850741915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,5120,0.10039733515845405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,3584,0.08900266885757446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,51200,0.9038747151692709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,4096,0.04115110966894362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,3584,0.03503200080659654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,16384,0.2972035672929552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,4096,0.05483022332191467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,3584,0.04994577831692166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,3072,0.04288444585270352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,3072,0.07706222269270155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,3072,0.03077955709563361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,3072,0.06342933575312297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,6144,0.05355733301904467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,2560,0.037179556157853864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,2048,0.051544000705083214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,2560,0.028516444895002577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,4096,0.08124533626768324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,2048,0.02358755634890662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,3584,0.07948088645935059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,2048,0.04683644572893778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,2560,0.06390844451056586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,1536,0.025480000509156123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,1536,0.020303999384244282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,1024,0.029787556992636785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,1536,0.03920355439186096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,2560,0.05481777919663323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,768,0.022631110416518316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,1024,0.020994666549894545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,768,0.014182221558358936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,2048,0.03237511052025689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,512,0.015541333291265698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,512,0.01351199961370892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,512,0.023728888895776536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,256,0.010403555300500657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,512,0.013040000365840064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,1024,0.01594044433699714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,256,0.012465777496496836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,128,0.008008889026112026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,1536,0.04052799940109253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,768,0.028020444843504164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,128,0.011399110986126794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,128,0.01144266708029641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,7168,0.13213688797420925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,128,0.016891555653678048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,32,0.008918222453859117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,64,0.011448000040319232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,32,0.015211555692884656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,256,0.011730666789743634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,64,0.008656000097592672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,65536,1.115240838792589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,768,0.016785777277416654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,51200,0.8684613439771863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,65536,0.4088924460940891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,65536,0.6340862380133735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,16384,0.2807359960344103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,16384,0.16083555751376682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,1024,0.032519999477598406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,51200,0.3229537804921468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,51200,0.4845564630296495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,16384,0.10790489117304485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,12288,0.12751111719343397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,12288,0.08636177910698785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,10240,0.16365777121649847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,16384,0.23949066797892252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,256,0.01934400035275353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,10240,0.10215555297003852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,8192,0.135862218009101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,12288,0.21326933966742623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,12288,0.18002843856811523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,7168,0.11654933293660481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,8192,0.08578577968809341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,10240,0.15755377875434026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,7168,0.07269689109590319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,8192,0.12242666880289714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,6144,0.1012035542064243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,10240,0.06765688790215386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,7168,0.11091466744740804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,7168,0.0505964457988739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,5120,0.08715999788708156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,8192,0.05536444319619072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,6144,0.09434933132595485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,5120,0.037420445018344455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,65536,0.9814400143093533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,5120,0.08082311020957099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,4096,0.04480977853139242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,3584,0.06236355834537082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,6144,0.04393511017163595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,4096,0.03359288970629374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,3584,0.04001511136690775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,3584,0.02924088968171014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,4096,0.07214311096403334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,3584,0.06041244665781657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,51200,0.7560879919264051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,2560,0.04686311218473646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,3072,0.02625333269437154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,3072,0.054007109668519765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,6144,0.06450311342875163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,2048,0.03828355669975281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,2560,0.030594666798909504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,4096,0.08297066556082831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,2560,0.04725689027044508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,5120,0.055325332615110606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,2048,0.019880000087949965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,2048,0.026800000005298193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,2048,0.04063733418782552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,1024,0.023245332969559565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,1536,0.02124533388349745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,3072,0.035106665558285184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,1536,0.03347377644644843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,1024,0.017364444004164804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,768,0.017231111725171406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,1024,0.028166221247778997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,1536,0.030188444587919448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,3072,0.05535199907090929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,512,0.012442666623327466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,768,0.012755555411179861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,768,0.024028443627887305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,1536,0.01641688909795549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,256,0.009359111388524374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,512,0.011709333293967776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,512,0.02051199972629547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,256,0.010675555302037133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,256,0.010383110907342698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,128,0.005987555616431766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,768,0.013817778064144982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,128,0.010058666268984476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,128,0.014653333359294467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,32,0.0052986666560173035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,512,0.011741333537631564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,64,0.00981155534585317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,32,0.013117333253224691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,2560,0.023176888624827068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,65536,0.33769599596659344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,256,0.017311111092567444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,65536,0.22860089937845865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,51200,0.2628577815161811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,64,0.005231110999981563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,65536,0.14448355303870306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,16384,0.0875377787484063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,51200,0.1116222275627984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,16384,0.061308443546295166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,16384,0.05762133333418104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,12288,0.06716799736022949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,16384,0.04043733411365085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,12288,0.046210666497548424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,128,0.01036177741156684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,10240,0.056780444251166455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,65536,0.20291733741760254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,10240,0.028090665737787884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,10240,0.03830400109291077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,8192,0.034605334202448525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,51200,0.17506399419572619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,51200,0.15654755963219538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,12288,0.04393777913517422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,7168,0.040538665321138166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,8192,0.024432889289326135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,10240,0.039847112364239164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,6144,0.03553422292073568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,1024,0.013808888693650564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,7168,0.029736889733208552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,6144,0.019327110714382596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,5120,0.030793779426150854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,12288,0.03215288784768846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,8192,0.03406577640109592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,5120,0.023062222533755835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,4096,0.025528000460730657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,5120,0.02358666724628872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,7168,0.021670222282409668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,8192,0.04801955487993029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,3584,0.022823111878501043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,4096,0.016136889656384785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,4096,0.019541333119074505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,6144,0.026538666751649644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,3072,0.020227554771635268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,3584,0.01758577757411533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,3584,0.019342222147517733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,3072,0.01550222271018558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,2560,0.017200888858901132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,3072,0.01757600075668759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,7168,0.029867556360032823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,2560,0.012368000215954252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,2048,0.014523555835088095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,2560,0.01588888963063558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,6144,0.0273297776778539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,2048,0.012513777448071374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,2048,0.011056889262464313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,2048,0.015013333823945789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,3072,0.013543999857372709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,5120,0.016837333639462788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,1536,0.010738666686746808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,2560,0.013838222457302941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,1024,0.009429333110650381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,1536,0.013092444174819522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,4096,0.020693333612547982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,768,0.008055110772450766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,3584,0.014416888356208801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,1024,0.012053333222866058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,768,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,1536,0.011708444191349877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,768,0.010437332921557956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,512,0.0069848886794514125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,512,0.007370666497283512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,1536,0.010103999740547603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,1024,0.009476444787449306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,1024,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,128,0.003944888710975647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,256,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,256,0.00871288859181934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,768,0.008166222108734978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,128,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,128,0.006330666856633292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,64,0.003759110967318217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,32,0.003900444342030419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,512,0.00923555592695872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,64,0.006614222294754452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,65536,0.2472595638699002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,32,0.008006222546100616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,512,0.006211555666393704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,65536,0.17557777298821342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,65536,0.12590221563975015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,51200,0.18850399388207328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,128,0.008385777473449707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,51200,0.1353786653942532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,51200,0.10081955459382798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,256,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,12288,0.05052711235152351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,16384,0.03692977627118429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,65536,0.16449422306484646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,16384,0.049110223849614464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,51200,0.12991555531819662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,12288,0.03670310974121094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,12288,0.03846755623817444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,256,0.004296888907750447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,8192,0.03533066643608941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,16384,0.0640266670121087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,10240,0.03373688790533278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,8192,0.02773688899146186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,16384,0.04810400141610039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,7168,0.030913776821560327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,8192,0.029674665795432195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,12288,0.02972266740269131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,10240,0.043143110142813794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,7168,0.02405244443151686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,7168,0.026549332671695288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,10240,0.03158488869667053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,5120,0.023973332511054143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,6144,0.024288000331984624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,8192,0.021749332547187805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,5120,0.018429333964983623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,4096,0.020510221521059673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,7168,0.01997066703107622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,6144,0.027266666293144226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,4096,0.0155395550860299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,3584,0.017880888448821176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,6144,0.021362667282422382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,4096,0.018950222267044913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,3584,0.013824889229403602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,3584,0.014377777775128683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,3072,0.015837333268589444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,3584,0.01773244473669264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,5120,0.020854221449957952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,3072,0.012465777496496836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,2560,0.013439111411571503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,3072,0.01626399987273746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,4096,0.014807111687130399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,2560,0.012000888586044312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,10240,0.025260445144441392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,2560,0.014997333288192749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,2048,0.010956444674068026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,1536,0.009689778089523315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,5120,0.016213332613309223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,3072,0.013148444394270578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,1536,0.009577777650621202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,1024,0.008237332933478886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,2048,0.011729777687125735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,1024,0.007976888782448238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,2048,0.010115555591053432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,1024,0.007932444413503012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,2048,0.014142221874660917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,768,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,768,0.009806222385830348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,512,0.004966222163703707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,1536,0.01297688898113039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,512,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,1024,0.010994666980372535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,512,0.007275555696752336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,256,0.004563555535342959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,256,0.005990222096443176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,1536,0.008742222355471717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,256,0.00832799987660514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,768,0.007368000017272101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,2560,0.011715555356608497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,128,0.00703111125363244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,128,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,64,0.0035742223262786865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,128,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,32,0.003567999849716822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,64,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,32,0.006959111326270633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,128,0.004063999901215236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,6144,0.018238221605618794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,51200,0.15874932871924505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,65536,0.12116000387403701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,768,0.0069857777820693115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,51200,0.12966844770643446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,16384,0.05493599838680691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,65536,0.15380088488260904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,256,0.006872889068391588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,16384,0.04605422086185879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,51200,0.12053778436448838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,12288,0.04298044575585259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,16384,0.03547022077772353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,512,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,16384,0.04689688815010918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,10240,0.03576711151334975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,65536,0.20471378167470297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,12288,0.028408000866572063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,12288,0.03528533379236857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,51200,0.0964755548371209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,12288,0.036778665251202054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,10240,0.024633778466118708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,65536,0.16676621966891816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,10240,0.03231288989384969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,8192,0.026467555099063452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,8192,0.02834933333926731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,7168,0.02277511192692651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,7168,0.025249777568711176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,8192,0.03088088830312093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,10240,0.030518223841985066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,6144,0.020242666204770405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,5120,0.020282665888468426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,6144,0.01719555589887831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,6144,0.023399111297395494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,4096,0.017380444539917838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,5120,0.015630222029156156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,7168,0.019080888893869188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,5120,0.020229332976871066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,3584,0.015101333459218344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,4096,0.014230221509933472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,4096,0.015202666322390238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,4096,0.01829866733815935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,3072,0.013175999952687157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,3584,0.017270222306251526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,8192,0.021132444341977436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,3072,0.013058666553762225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,3072,0.012192000117566852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,2560,0.011422221859296164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,5120,0.01737955543729994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,3072,0.01629244453377194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,2048,0.009927111367384592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,6144,0.023042667243215773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,2560,0.011824000212881299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,2560,0.01497511068979899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,1536,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,3584,0.014164444473054675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,2048,0.01388533330625958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,7168,0.02606577840116289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,1024,0.007098666495747036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,1536,0.009200000100665623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,1536,0.0124195557501581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,1024,0.008000888758235509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,2560,0.010565333068370819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,1024,0.010440888504187265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,2048,0.009405333134863112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,768,0.006809777683681912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,512,0.004586666822433472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,768,0.009832888841629028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,768,0.007663110891977946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,1536,0.00870311094654931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,256,0.003922666526503033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,512,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,512,0.008922666311264038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,1024,0.007771555748250749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,768,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,256,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,2048,0.01070666644308302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,128,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,64,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,128,0.006665777828958299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,128,0.007673777639865875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,512,0.007167110840479533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,64,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,32,0.00658044425977601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,65536,0.1814346710840861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,256,0.005994666781690385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,65536,0.12691733572218153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,51200,0.1421626673804389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,3584,0.01315822203954061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,32,0.0038782221575578055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,51200,0.05817333194944593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,16384,0.0495013329717848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,65536,0.14079822434319392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,16384,0.03502666619088914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,128,0.003567999849716822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,16384,0.023194667365815904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,16384,0.04290666513972812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,65536,0.07181600067350599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,12288,0.01872177753183577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,10240,0.03216444452603658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,12288,0.03435733252101474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,10240,0.023499555057949487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,10240,0.016569778323173523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,8192,0.027727999620967444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,10240,0.03005689051416185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,256,0.006761777732107375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,8192,0.014799111419253878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,7168,0.024139554964171514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,8192,0.02607111136118571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,51200,0.09601866536670262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,7168,0.017788444956143696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,7168,0.013663999736309052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,51200,0.11098311344782512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,12288,0.038980444272359215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,6144,0.01276622215906779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,8192,0.02051733268631829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,6144,0.02204977803760105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,5120,0.014120888378885059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,6144,0.020966221888860066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,4096,0.015221332510312399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,5120,0.011744889120260874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,6144,0.016196444630622864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,4096,0.012845333251688214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,3584,0.014127110441525778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,12288,0.027037334111001756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,3584,0.011726222104496427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,7168,0.02385422256257799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,5120,0.019415999452273052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,3072,0.010932444698280759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,3072,0.015458666616015963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,4096,0.010986666712496014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,5120,0.017893332574102614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,2560,0.009961778091059791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,3584,0.01645599967903561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,3072,0.012636444634861417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,2560,0.008353778057628209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,2048,0.009194666312800514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,2048,0.013194666968451606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,4096,0.017308443784713745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,1536,0.008236444658703273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,2048,0.00774133370982276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,1536,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,1536,0.007777777810891469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,1536,0.011794666449228922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,3584,0.009705777797434065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,1024,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,1024,0.0070053330726093715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,768,0.004951111144489712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,1024,0.00942666663063897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,3072,0.009010666774378883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,768,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,768,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,768,0.009033777647548253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,512,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,256,0.0038515557017591265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,512,0.00869599978129069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,2048,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,256,0.006052444378534953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,256,0.005617777920431561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,128,0.0038782221575578055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,256,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,2560,0.01109155515829722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,512,0.004560889055331548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,64,0.003579555700222651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,32,0.003587555554178026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,1024,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,64,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,32,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,65536,0.12119555473327637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,128,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,128,0.005568000177542369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,65536,0.06890933381186591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,128,0.008006222546100616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,512,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,65536,0.138044450018141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,2560,0.014149333039919535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,16384,0.04939111073811849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,51200,0.05617688761817085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,65536,0.1627697812186347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,12288,0.03597955571280585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,16384,0.022466666168636743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,16384,0.034197333786222674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,51200,0.11005155245463054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,12288,0.0181377778450648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,12288,0.026377777258555096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,12288,0.03422933485772874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,8192,0.026053332620196875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,10240,0.01609155535697937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,51200,0.12671732902526855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,10240,0.029313776228162978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,7168,0.022330666581789654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,8192,0.019708444674809773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,51200,0.09467289182874893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,8192,0.02586577832698822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,6144,0.01960000064637926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,7168,0.017419555120997958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,10240,0.029751112063725788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,6144,0.015842666228612263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,6144,0.012430222498046027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,6144,0.0214035560687383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,8192,0.01442311041884952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,5120,0.014147554834683737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,5120,0.01919911139541202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,4096,0.014559111661381192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,5120,0.011411555939250521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,7168,0.013403555585278405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,3584,0.012159999873903064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,4096,0.010038221875826517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,4096,0.012825777961148156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,5120,0.01687733332316081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,10240,0.02273955610063341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,3072,0.011199110911952125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,3584,0.009136888715955947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,3072,0.010727110836240979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,3072,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,16384,0.042295998997158475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,4096,0.017346666918860543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,3584,0.011591110792424945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,3584,0.0158942225906584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,2560,0.013889777991506787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,2048,0.00904177791542477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,7168,0.02346399923165639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,2048,0.007442666424645319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,1536,0.007384888827800751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,2048,0.013040888640615674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,2560,0.008011555506123437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,2048,0.008652444514963362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,1536,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,1536,0.006960888703664144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,3072,0.01517777807182736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,1024,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,768,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,1024,0.009372444616423713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,768,0.0064026667839950984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,768,0.0063484443558586975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,768,0.009383999639087254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,1024,0.005988444305128521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,512,0.006004444426960415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,1536,0.011077333655622272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,256,0.003941333128346337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,512,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,1024,0.006759111252095964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,128,0.003540444291300244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,256,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,256,0.008010666403505538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,2560,0.009885333478450775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,512,0.004578666554556953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,128,0.0053119998839166425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,32,0.0036728887094391715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,64,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,512,0.008656889200210571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,32,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,65536,0.13413688871595594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,2560,0.009720889230569204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,256,0.006112888869312074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,51200,0.10797244310379028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,65536,0.11493333180745442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,65536,0.06647555695639716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,16384,0.04292088747024536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,128,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,51200,0.08966755867004395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,128,0.007852444218264686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,65536,0.13773777749803331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,12288,0.03533422284656101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,16384,0.02123111155298021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,16384,0.03296533226966858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,51200,0.10966222153769599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,10240,0.024530667397711012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,16384,0.0425182216697269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,12288,0.01754399968518151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,12288,0.025436444414986506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,12288,0.03386577632692125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,10240,0.01564711166752709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,10240,0.021577777134047613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,7168,0.0181351105372111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,8192,0.0189057770702574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,64,0.003540444291300244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,8192,0.025198222862349615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,51200,0.05400800042682224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,7168,0.016944888565275405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,7168,0.02330488959948222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,6144,0.015499555402331881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,8192,0.021176889538764954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,10240,0.029231998655531142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,8192,0.01390044391155243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,5120,0.010823110739390055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,7168,0.012990222209029727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,4096,0.01240799989965227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,5120,0.018617777360810172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,4096,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,6144,0.012069332930776807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,6144,0.021588444709777832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,4096,0.016884444488419425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,3584,0.009008888569143083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,5120,0.01367911116944419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,3584,0.015817777978049386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,6144,0.015916445189052157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,3072,0.010647999743620554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,3072,0.008206222620275285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,3072,0.015100444356600443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,4096,0.012086222569147745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,5120,0.01386577801571952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,2048,0.007965332931942409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,2560,0.007778666913509369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,2560,0.013902222116788229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,3584,0.01148888882663515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,3072,0.009866666462686326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,1536,0.0069439998931354946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,2048,0.007350222104125553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,2048,0.008001777860853408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,2048,0.01220444424284829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,1024,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,1536,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,2560,0.008936000367005667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,1536,0.010728889041476779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,768,0.004632888982693355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,3584,0.011054221954610614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,1024,0.00666933341158761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,1024,0.009715555442704095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,768,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,768,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,768,0.008755555583371056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,512,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,256,0.003535110917356279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,512,0.00868266655339135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,1536,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,1024,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,256,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,256,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,2560,0.009375111096435124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,512,0.00423555572827657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,64,0.0028702221396896574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,128,0.00599911105301645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,128,0.0052471111218134565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,32,0.003244444520937072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,128,0.008001777860853408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,512,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,64,0.005611555443869696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,65536,0.11521422863006592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,32,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,51200,0.09352266788482666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,65536,0.10979466968112522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,65536,0.06415910853279962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,16384,0.03836177786191305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,51200,0.08614399698045518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,51200,0.051784889565573804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,65536,0.13708178202311197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,16384,0.020432889461517334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,51200,0.10946133401658799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,16384,0.04186755418777466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,12288,0.02409066590997908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,12288,0.016958221793174744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,12288,0.03296444482273526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,10240,0.02104355560408698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,128,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,12288,0.030403554439544678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,8192,0.01868977811601427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,8192,0.013442666994200813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,10240,0.026921777261628047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,7168,0.016726222303178575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,8192,0.02488800055450863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,7168,0.012440889245933957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,10240,0.015205333630243937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,6144,0.014466666513019137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,8192,0.0188071115149392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,7168,0.02295644415749444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,5120,0.012744888663291931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,6144,0.011476444701353708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,6144,0.015097777048746744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,6144,0.021018667353524104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,4096,0.01183555523554484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,5120,0.01349244432316886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,5120,0.01862133377128177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,4096,0.012090666426552666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,4096,0.008975110948085785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,3584,0.010087110930018956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,7168,0.01657511128319634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,256,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,3072,0.009343999955389235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,3584,0.008340444829728868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,3584,0.015511110424995422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,16384,0.03141866789923774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,3072,0.01014311114947001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,3072,0.007978666987684038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,3072,0.01483022173245748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,2560,0.00868800034125646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,2560,0.007669332954618666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,2048,0.007641777396202087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,10240,0.02900444467862447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,3584,0.010811555716726514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,2048,0.0073520003093613525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,1536,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,2560,0.00832444429397583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,1536,0.007332444190979004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,1536,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,2560,0.013763555222087435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,1024,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,1024,0.006646222124497096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,768,0.004489777816666497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,1024,0.009717333647939894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,5120,0.010392000277837118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,768,0.006183111005359226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,768,0.00905955582857132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,4096,0.0166302224000295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,512,0.0057653333577844836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,2048,0.008120888637171851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,512,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,256,0.0035946667194366455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,1024,0.005232000102599462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,256,0.008018666671382057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,256,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,128,0.0035137778355015647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,768,0.006618666566080517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,128,0.007358222372002072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,2048,0.012422222230169507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,64,0.0029075555503368378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,1536,0.010420444111029306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,64,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,32,0.0032239999208185407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,65536,0.09255644347932603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,32,0.005716444303592046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,65536,0.06914222240447998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,512,0.008345777789751688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,65536,0.061716443962521024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,65536,0.13727022541893855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,51200,0.05351288782225715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,256,0.00544177782204416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,51200,0.04988444513744778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,51200,0.1091537806722853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,16384,0.019866666860050626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,12288,0.02574311031235589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,128,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,12288,0.016581333345837064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,12288,0.01648000048266517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,10240,0.022988445229000513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,16384,0.021016000045670405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,10240,0.014746665954589844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,51200,0.07632444302241008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,10240,0.014582221706708273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,8192,0.015962666935390897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,10240,0.02866577770974901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,8192,0.013219555219014486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,512,0.004053333153327306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,7168,0.014456000592973499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,7168,0.012109333442317115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,16384,0.040955556763543025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,7168,0.02273422148492601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,6144,0.012920888761679331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,7168,0.011808888779746162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,6144,0.010780444575680627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,6144,0.011406222151385413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,5120,0.011499555574523078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,6144,0.020949333906173706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,8192,0.012772444221708508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,128,0.005316444569163852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,4096,0.010875555376211802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,5120,0.010080888867378235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,4096,0.00867733359336853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,5120,0.018527110417683918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,3584,0.00906044410334693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,4096,0.01665422154797448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,3584,0.015645333462291293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,3072,0.00813422186507119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,3584,0.008107555409272512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,12288,0.03295288814438714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,8192,0.024833778540293377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,2560,0.007584888901975419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,3072,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,4096,0.009272000028027428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,3072,0.014189332723617554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,2048,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,2560,0.007689778175618913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,2560,0.012492444780137805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,2048,0.006514666808976068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,1536,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,3584,0.008823110825485652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,2048,0.011779555843936073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,16384,0.0325564444065094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,5120,0.009376889301670922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,1536,0.006991110742092133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,3072,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,1024,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,2560,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,1536,0.010433778166770935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,1024,0.009708444277445475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,768,0.006312888943486744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,2048,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,768,0.009073778159088558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,1024,0.004281777888536453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,512,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,256,0.003552000141806073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,512,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,1536,0.0063786668082078295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,768,0.0038862224254343244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,128,0.0031911111954185697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,256,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,256,0.007975999679830339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,768,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,128,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,128,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,32,0.0031946665710873077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,64,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,512,0.008720888859695857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,65536,0.08485955662197536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,1024,0.00664533343580034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,256,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,65536,0.06076888905631172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,51200,0.06822666856977674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,512,0.0037964445849259696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,65536,0.13673067092895508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,51200,0.05227022369702657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,51200,0.04894577794604831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,51200,0.10918400022718643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,32,0.005588444570700328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,16384,0.020041777027977836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,16384,0.019880889190567862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,65536,0.067358222272661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,16384,0.040898667441474065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,128,0.00508266646001074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,12288,0.016237333416938782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,10240,0.021609778205553692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,12288,0.032969776127073504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,16384,0.029690666331185236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,8192,0.015336889359686108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,10240,0.014523555835088095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,10240,0.028394665982988145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,12288,0.02440622283352746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,7168,0.013771555489963956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,8192,0.013176889055305056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,64,0.003198222153716617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,12288,0.016227554943826463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,6144,0.012086222569147745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,7168,0.011387555963463254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,10240,0.014194667339324951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,7168,0.022678222921159532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,5120,0.01086400035354826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,6144,0.010065777434243096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,6144,0.010882666541470421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,6144,0.02095555596881443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,5120,0.00904266701804267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,5120,0.009714666340086196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,5120,0.018237334158685472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,3584,0.008715555899673039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,4096,0.008739555875460306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,4096,0.008776889079146916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,4096,0.016346666547987197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,3072,0.008557333714432186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,7168,0.011763555308183035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,3584,0.015231110983424716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,3072,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,2560,0.007368889119890001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,3072,0.013441777891582914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,8192,0.012738666600651212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,2560,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,2048,0.006244444598754247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,2560,0.012421333127551608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,3584,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,2048,0.006527110934257507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,1536,0.0052355556852287715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,2048,0.011402666568756104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,3072,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,4096,0.010048888623714447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,1536,0.006487999939256244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,1024,0.004451555510361989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,1536,0.010391111175219217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,2560,0.006675555474228329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,768,0.0038933331767717996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,1024,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,1024,0.009396444592210982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,8192,0.02514133354028066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,512,0.003549333247873518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,768,0.006331555545330048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,768,0.00869511150651508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,512,0.0053182223604785065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,256,0.003528888854715559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,512,0.008383110993438298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,1536,0.0069395556218094295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,1024,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,256,0.006033777776691649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,256,0.008041777544551426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,3584,0.007977777885066139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,768,0.005608888963858287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,64,0.0029724445194005966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,128,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,128,0.005111999809741974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,128,0.007672000262472365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,512,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,65536,0.0740728908114963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,32,0.005499555418888728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,65536,0.06315466430452135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,65536,0.05959022045135498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,256,0.0053137776752312975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,65536,0.13719111018710667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,2048,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,16384,0.02511022157139248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,51200,0.10859644412994385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,16384,0.019317333896954853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,16384,0.019354666272799175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,51200,0.058768888314565025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,32,0.0032097777972618737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,64,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,12288,0.032448000378078885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,10240,0.0184515549076928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,128,0.003544888976547453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,10240,0.01385155568520228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,51200,0.047397332059012524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,10240,0.0281368891398112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,8192,0.01629155543115404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,12288,0.016026667422718473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,8192,0.012439111040698158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,12288,0.0214035560687383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,8192,0.012240889171759287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,7168,0.014780445231331719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,8192,0.024280889166726008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,51200,0.05052622159322103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,10240,0.01424888935354021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,6144,0.011724444727102915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,7168,0.02243288854757945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,6144,0.01016711112525728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,6144,0.010493333141009012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,5120,0.010835555692513784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,6144,0.020600888464185927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,16384,0.04080177678002252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,4096,0.009321778184837764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,5120,0.01825777855184343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,12288,0.01606044504377577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,4096,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,3584,0.008406221866607666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,4096,0.008674666285514832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,3584,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,3072,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,3584,0.014534221755133735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,3584,0.008170666793982187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,3072,0.008004444340864817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,3072,0.007349333001507654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,2560,0.007294221884674496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,7168,0.01092266705301073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,2560,0.007428444094128079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,2560,0.006656000183688269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,2560,0.012434666355450949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,2048,0.006610666712125142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,5120,0.009368889033794403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,2048,0.011418666276666852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,2048,0.007258666886223688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,7168,0.011426666544543372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,1536,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,1024,0.004872888740566042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,4096,0.01629066632853614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,1536,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,1024,0.0068408888247278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,1024,0.00564444437623024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,1024,0.009351111120647853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,5120,0.009334222310119206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,1536,0.0052782222628593445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,512,0.003903111235962974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,768,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,768,0.009355555805895064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,3072,0.013280889226330651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,256,0.0035537779331207275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,512,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,512,0.008340444829728868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,256,0.005962666538026597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,768,0.004286222159862518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,256,0.007816889219813878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,768,0.005616888817813661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,128,0.007674666742483775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,128,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,1536,0.010741333166758219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,32,0.00318755561278926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,64,0.0052782222628593445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,256,0.0052675555149714155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,65536,0.06337777773539226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,32,0.005314666777849197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,65536,0.06124533547295464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,65536,0.058541331026289195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,51200,0.05234400100178189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,65536,0.1367608838611179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,64,0.003026666740576426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,16384,0.02199111051029629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,512,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,51200,0.04850577645831638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,51200,0.10861244466569688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,16384,0.01900800069173177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,12288,0.0189057770702574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,16384,0.04081066780620151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,12288,0.015443555182880826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,128,0.004962666581074397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,10240,0.01625955601533254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,10240,0.013471999930010902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,51200,0.04637955625851949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,10240,0.028184887435701158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,10240,0.013849777479966482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,16384,0.01867555578549703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,8192,0.012243555651770698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,7168,0.013291555146376291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,8192,0.024689777029885188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,12288,0.032496889432271324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,7168,0.01073333372672399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,7168,0.011433777709801992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,6144,0.01276177747382058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,7168,0.02236000034544203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,8192,0.0151573336786694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,5120,0.01016533292002148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,6144,0.010417777631017896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,6144,0.01038577738735411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,6144,0.020615999897321064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,4096,0.008994667066468133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,5120,0.009088888764381409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,5120,0.009072889056470659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,12288,0.015834665960735746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,4096,0.00850933293501536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,3584,0.008353778057628209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,4096,0.015203555425008139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,2048,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,3584,0.008018666671382057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,128,0.0032222221295038858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,3072,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,8192,0.011243555280897351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,5120,0.018182223041852314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,3072,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,3072,0.01311555587583118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,4096,0.008031110796663497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,2560,0.00738666703303655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,2048,0.00564444437623024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,3584,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,3584,0.014409777190950183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,2048,0.006612444503439798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,1536,0.005025777965784073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,3072,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,2560,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,1536,0.005992888990375731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,1024,0.004546666724814309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,2560,0.006994666324721442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,2560,0.012432888978057437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,768,0.003928889003064898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,1024,0.006501333167155583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,1024,0.005647111270162795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,1024,0.009364444348547194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,512,0.003580444388919406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,2048,0.011410666836632622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,512,0.0052915554907586836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,1536,0.010041777458455827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,512,0.00867733359336853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,256,0.003543111185232798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,768,0.006303999986913469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,256,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,768,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,128,0.003212444484233856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,256,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,1536,0.007010666860474481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,64,0.002879111096262932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,128,0.007582222421964009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,32,0.0028773333049482773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,512,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,64,0.0052915554907586836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,32,0.005247999810510212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,65536,0.058261334896087646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,65536,0.05671911107169258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,65536,0.1367822223239475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,51200,0.04398933384153578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,65536,0.053695112466812134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,51200,0.046463999483320445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,16384,0.01889155473974016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,2048,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,51200,0.10871289173762004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,16384,0.018718222777048748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,16384,0.018239999810854595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,12288,0.015465777781274585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,16384,0.04052266809675429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,12288,0.015016888578732809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,12288,0.015082667271296183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,10240,0.013811555173661975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,12288,0.03227644496493869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,10240,0.01315822203954061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,10240,0.013388444152143268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,51200,0.04521955715285408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,10240,0.028104000621371802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,8192,0.011167111496130625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,8192,0.011402666568756104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,7168,0.012084444363911947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,128,0.006011555592219035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,7168,0.010597333312034607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,7168,0.01072799993885888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,7168,0.02204888893498315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,768,0.005580444302823808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,8192,0.013814222481515674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,5120,0.010044444766309526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,6144,0.020276443825827706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,6144,0.009708444277445475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,5120,0.00902133352226681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,8192,0.02427022159099579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,4096,0.008653333617581261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,5120,0.016979555288950603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,6144,0.01125955581665039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,4096,0.008387555678685507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,4096,0.01516355574131012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,256,0.005044444567627377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,3584,0.007637333538797166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,3584,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,3584,0.014153778553009033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,128,0.00490488898422983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,5120,0.008700444466537898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,3072,0.007024889190991719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,3072,0.007826666865083907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,3072,0.013111111190583972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,3584,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,2048,0.005573333137565189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,2560,0.006693333387374878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,6144,0.009359111388524374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,3072,0.007279111279381647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,2048,0.007154666715198093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,1536,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,2560,0.0063182223174307085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,4096,0.00796088907453749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,1536,0.006103111224042044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,1024,0.0042275554604000514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,1536,0.006649777707126405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,2560,0.012273777690198688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,2048,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,768,0.003871110992299186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,1024,0.009724444813198512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,2048,0.011060444845093621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,768,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,768,0.0053680001033677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,512,0.0035342222286595237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,768,0.009015999734401703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,2560,0.007335111498832703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,512,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,1536,0.01038755559258991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,512,0.0083137775460879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,1024,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,256,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,256,0.007681777907742395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,128,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,128,0.004951111144489712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,128,0.008050666915045844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,32,0.0028808888875775864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,64,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,512,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,65536,0.041787554820378624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,32,0.0052977779673205475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,1024,0.006607111129495833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,256,0.005223111146026188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,51200,0.03348533312479655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,65536,0.05589510997136434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,65536,0.13673778374989828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,16384,0.014803555276658801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,51200,0.04463288850254483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,51200,0.10831377903620402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,256,0.003216000066863166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,16384,0.017705778280893963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,12288,0.014217777384652032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,16384,0.040433777703179255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,128,0.0028942223224374983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,65536,0.056187556849585645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,12288,0.03222577770551046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,64,0.0028524444335036804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,51200,0.04491110973887973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,10240,0.012822222378518848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,8192,0.01166488892502255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,10240,0.012471111284361945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,16384,0.017934223016103108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,8192,0.01111822244193819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,8192,0.02404977712366316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,12288,0.01386755539311303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,12288,0.014644443988800049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,6144,0.010594666832023196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,7168,0.010394666757848527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,7168,0.021321778496106465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,6144,0.009901333186361525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,5120,0.009885333478450775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,6144,0.01947733263174693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,7168,0.011110222174061669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,5120,0.008647999829716152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,5120,0.008999110923873054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,4096,0.008412444757090675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,5120,0.01726222203837501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,10240,0.012773333324326409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,3584,0.007637333538797166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,4096,0.008340444829728868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,4096,0.007719111111429002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,4096,0.015217777755525378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,3072,0.006648889018429651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,6144,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,10240,0.027660444378852844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,3584,0.014149333039919535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,2560,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,8192,0.010779555473062726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,3072,0.007009777757856581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,3072,0.01310755560795466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,2048,0.005247999810510212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,2560,0.007706666986147563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,2560,0.00665866666369968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,2560,0.0124177775449223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,1536,0.004906666775544484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,2048,0.007032000356250339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,2048,0.011431111229790581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,3584,0.008169777691364288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,3584,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,1024,0.004229333251714706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,1536,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,1536,0.010739555789364709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,3072,0.007716444631417592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,768,0.0038497779104444715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,1024,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,1024,0.00924177798959944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,512,0.003549333247873518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,768,0.005568000177542369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,768,0.00902133352226681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,2048,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,512,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,512,0.005998222364319696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,256,0.003375111147761345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,512,0.008674666285514832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,256,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,256,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,256,0.007660444411966536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,768,0.006267555471923616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,7168,0.010430222584141625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,64,0.0028862222615215513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,128,0.005909333212508096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,32,0.0028764444092909494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,128,0.007661333514584436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,65536,0.031826668315463595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,64,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,1536,0.00699733363257514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,32,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,1024,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,65536,0.05537866552670797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,65536,0.13661332925160727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,51200,0.042972445487976074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,128,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,16384,0.012866666747464074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,51200,0.10813333590825398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,16384,0.016233777006467182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,16384,0.01759022143152025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,16384,0.040256887674331665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,51200,0.025799110531806946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,128,0.0032479998966058097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,12288,0.014006222287813822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,12288,0.031553778383466936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,10240,0.011980444192886353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,51200,0.0442737771405114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,8192,0.01163200040658315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,12288,0.013844444519943662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,8192,0.010699555277824402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,7168,0.01071733319097095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,8192,0.023021333747439917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,8192,0.01107022249036365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,65536,0.054400000307295054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,6144,0.010836444795131683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,7168,0.010100444157918295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,7168,0.021242666575643752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,10240,0.012220444778601328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,6144,0.00925511121749878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,6144,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,6144,0.019334221879641216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,5120,0.00905688852071762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,4096,0.00794933322403166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,12288,0.01351111133893331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,7168,0.010420444111029306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,4096,0.008009778128729926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,4096,0.008376888930797577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,4096,0.015150222513410779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,10240,0.02714666724205017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,3584,0.007335111498832703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,3072,0.006588444527652528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,5120,0.008467555873923832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,3584,0.014144000079896716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,5120,0.016890666551060148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,2560,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,3072,0.007763555480374231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,3072,0.006983111302057902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,3072,0.012867555850081973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,3584,0.007345777418878343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,2560,0.0074177781740824384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,5120,0.009341333475377824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,2048,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,1536,0.004896000027656555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,2048,0.01125511113140318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,10240,0.01276711126168569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,1536,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,1536,0.010144888526863521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,1536,0.006896889044178857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,2048,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,2560,0.012120000190205045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,1024,0.005630222045713001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,1024,0.0063075555695427795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,768,0.005459555735190709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,768,0.008752000000741746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,768,0.0063279999627007385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,512,0.0033617777129014335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,1024,0.004221333397759331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,512,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,2560,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,512,0.008346666892369589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,1024,0.009392889009581672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,256,0.0052675555149714155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,256,0.005974222388532426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,128,0.0031999999450312722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,128,0.004960000101062986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,128,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,64,0.002867555452717675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,512,0.005290666802061929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,256,0.0032097777972618737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,768,0.003952888978852166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,768,65536,0.026870222555266485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,2048,0.0063182223174307085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,64,0.004934222333961063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,32,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,256,0.007624000310897827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,768,51200,0.021998221675554912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,65536,0.05312800076272753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,65536,0.1363555590311686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,51200,0.04211644331614176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,51200,0.044065776798460216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,32,0.0028942223224374983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,3584,0.008070222205585903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,16384,0.01660977800687154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,768,12288,0.011729777687125735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,16384,0.04011022382312351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,65536,0.05480622251828512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,12288,0.013471110827393003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,12288,0.013458666702111563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,128,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,768,16384,0.013850666582584381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,51200,0.1080906656053331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,768,8192,0.009846222069528367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,10240,0.011857777833938599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,10240,0.026779555612140234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,16384,0.016192888220151264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,8192,0.010773333410422007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,8192,0.011078221930397881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,8192,0.022679110368092854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,768,6144,0.00884622252649731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,7168,0.010390222072601318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,7168,0.010034666293197209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,7168,0.021200888686709937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,768,5120,0.008003555238246918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,6144,0.009117333425415887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,6144,0.01899111105336083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,5120,0.008395555946562026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,768,4096,0.007975111405054728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,768,7168,0.009374221993817223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,4096,0.007631999750932057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,768,10240,0.011211555865075854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,4096,0.014808000789748298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,4096,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,6144,0.009736888938479954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,10240,0.012279111478063794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,768,3072,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,3584,0.014409777190950183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,5120,0.016880000631014507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,3072,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,768,2560,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,3072,0.01311644415060679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,12288,0.03124177787039015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,768,3584,0.007299555672539606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,768,2048,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,2560,0.007333333293596904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,2560,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,2560,0.012409778104888068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,3584,0.008074666890833113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,2048,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,2048,0.006312888943486744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,2048,0.011395555403497485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,768,1024,0.0039057777159743835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,1536,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,3072,0.0070257774657673305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,1024,0.0063075555695427795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,768,768,0.003539555602603488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,3584,0.0073520003093613525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,768,0.005562666803598404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,5120,0.008998221821255153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,768,512,0.003546666767862108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,768,0.0063075555695427795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,1536,0.005988444305128521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,768,0.008976000050703684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,768,256,0.002862222285734283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,512,0.00600622221827507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,512,0.0052728888889153796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,1024,0.005655111124118169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,512,0.008349333372380998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,768,128,0.0029004443850782183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,256,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,256,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,768,64,0.0025680000997251934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,128,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,128,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,128,0.008027555214034187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,512,65536,0.020981333321995203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,64,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,32,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,1536,0.01038577738735411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,512,51200,0.01646222174167633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,65536,0.05185955431726244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,1024,0.009706666900051964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,51200,0.04180622100830078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,51200,0.039622220728132464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,768,1536,0.004583999928500918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,16384,0.01620533400111728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,16384,0.03911555475658841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,16384,0.016125332978036668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,65536,0.05390577846103244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,65536,0.13569511307610407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,12288,0.013467555244763693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,256,0.007807999849319458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,512,10240,0.010047111246320937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,12288,0.030530667967266504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,768,32,0.0028737777223189673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,10240,0.012319111161761813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,10240,0.011766222616036734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,512,8192,0.00870577742656072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,10240,0.02643377747800615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,512,12288,0.009696000152164036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,512,7168,0.009477333062224919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,8192,0.010495999621020423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,8192,0.022671999202834234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,12288,0.013455111119482251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,7168,0.010377777947319878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,512,6144,0.009008000294367472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,7168,0.021035555336210463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,6144,0.009744889206356471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,6144,0.009006222089131674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,6144,0.019270223047998216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,8192,0.011074666347768573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,5120,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,512,4096,0.00741511086622874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,5120,0.01688088807794783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,7168,0.009692444569534725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,4096,0.00833422193924586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,512,3584,0.0069848886794514125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,4096,0.01519733336236742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,512,5120,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,512,3072,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,3072,0.0070053330726093715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,512,16384,0.01089511149459415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,51200,0.10815022389094035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,512,2560,0.0058906666106647914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,3584,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,3584,0.014293332894643148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,2560,0.00666933341158761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,2560,0.007344000041484833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,5120,0.008917333351241218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,2048,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,4096,0.007631999750932057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,512,1536,0.004563555535342959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,3072,0.013394667042626275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,3584,0.008151111503442129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,1536,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,1536,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,1536,0.010436444646782346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,512,2048,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,2560,0.012337778177526263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,1024,0.005631111148330901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,512,768,0.003559111307064692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,1024,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,3072,0.007928888830873702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,768,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,512,512,0.0032097777972618737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,768,0.009011555876996782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,512,1024,0.0041928887367248535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,512,256,0.0032213332338465583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,1024,0.006665777828958299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,256,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,256,0.00563822231358952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,2048,0.01108977778090371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,256,0.00813333359029558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,128,0.004944889081848992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,512,0.008594666918118795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,512,64,0.002551111082235972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,64,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,512,32,0.002519999941190084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,32,0.004963555683692296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,256,65536,0.016175111134847004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,128,0.007977777885066139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,128,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,65536,0.0473235547542572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,256,51200,0.018564444449212816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,65536,0.13532177607218424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,2048,0.007279111279381647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,51200,0.038041777080959745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,512,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,256,16384,0.010032888915803697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,51200,0.10769689083099365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,16384,0.0158915552828047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,65536,0.05106400118933784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,16384,0.03915999995337592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,768,0.005595555735958948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,12288,0.013441777891582914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,256,10240,0.009349333743254343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,12288,0.030826667944590252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,512,0.0052791109515561005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,10240,0.011742221812407175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,256,8192,0.008336000144481659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,16384,0.01622488929165734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,256,12288,0.010054222411579555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,8192,0.010370666782061258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,256,7168,0.008089777496125963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,12288,0.013528889252079858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,7168,0.009666666388511658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,10240,0.012086222569147745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,7168,0.021320000290870667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,256,6144,0.00871733327706655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,6144,0.009295110901196798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,8192,0.010776888993051318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,6144,0.019144000278578866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,256,5120,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,6144,0.009738666315873465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,51200,0.03949422306484646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,5120,0.008978666530715095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,5120,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,256,4096,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,7168,0.010403555300500657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,4096,0.007650666766696506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,4096,0.008438222110271454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,4096,0.014840889308187695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,8192,0.023050665855407715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,3584,0.007326222128338284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,256,3072,0.006256888724035687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,3584,0.008051555189821456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,3584,0.013807111316257052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,5120,0.01699733402993944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,256,2560,0.005766222046481238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,3072,0.01314311143424776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,256,3584,0.006940444310506185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,512,128,0.002903999967707528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,2560,0.007351111206743453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,256,2048,0.004910222358173794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,2560,0.012133333418104382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,2048,0.006994666324721442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,2048,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,2048,0.011008888483047485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,3072,0.007666666474607255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,10240,0.026777777406904433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,256,1024,0.0038204445607132385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,1536,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,1536,0.010049777726332346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,256,768,0.003539555602603488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,2560,0.006664888726340399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,1024,0.009699555734793345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,768,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,768,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,256,512,0.003232888877391815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,768,0.00905955582857132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,3072,0.00702133360836241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,512,0.005984888722499211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,256,256,0.0028977776981062363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,1536,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,256,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,256,128,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,256,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,128,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,256,1536,0.004556444370084339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,256,64,0.0028826666788922418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,128,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,256,32,0.0028417776856157514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,512,0.008011555506123437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,64,0.004972444640265571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,1024,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,32,0.004906666775544484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,1024,0.005578666511509154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,128,51200,0.013456888496875763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,65536,0.046559111939536206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,65536,0.04911466770701938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,65536,0.13364444838629827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,128,0.007967111137178209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,128,16384,0.009714666340086196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,51200,0.10603377554151748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,16384,0.016174222032229107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,16384,0.015984000431166757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,16384,0.038732445902294584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,256,0.005249777601824866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,12288,0.01351999988158544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,12288,0.013295110729005603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,128,10240,0.00905777762333552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,12288,0.03087377879354689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,51200,0.03896266553137038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,51200,0.03775022096104092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,10240,0.012120000190205045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,512,0.0052933332820733385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,8192,0.010437332921557956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,128,7168,0.007819555699825287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,7168,0.009787555370065901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,128,8192,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,10240,0.02679644525051117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,7168,0.020964443683624268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,7168,0.01039644413524204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,8192,0.011065777805116443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,8192,0.022679999470710754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,128,5120,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,6144,0.019253333409627277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,5120,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,10240,0.011741333537631564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,5120,0.0090515555606948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,128,6144,0.007348444726732042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,128,65536,0.0151582227812873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,4096,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,128,3584,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,4096,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,4096,0.015156444576051501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,3584,0.008048888709810045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,3584,0.007266666326257918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,3584,0.013960000541475085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,5120,0.016872000363137987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,128,4096,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,128,2560,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,3072,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,3072,0.007639110916190677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,3072,0.012799111505349478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,6144,0.009702222214804756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,128,2048,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,128,3072,0.006695111178689533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,2048,0.006304889089531369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,2048,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,128,1536,0.004581333448489507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,2048,0.0114720000161065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,6144,0.009089777866999308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,1536,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,128,1024,0.0038924444880750445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,1024,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,128,12288,0.009720889230569204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,1024,0.009487999810112847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,1024,0.006308444258239534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,128,768,0.003504000190231535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,2560,0.01238666640387641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,128,512,0.0031902222997612427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,768,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,768,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,512,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,128,256,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,512,0.008635555704434713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,256,0.005071111023426056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,2560,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,256,0.008045333127180735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,256,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,1536,0.010423111418883005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,128,0.006160000132189856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,128,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,128,64,0.002603555512097147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,128,32,0.002568888788421949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,64,0.005057777795526716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,64,65536,0.014662222729788886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,2560,0.006607999818192587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,128,128,0.0029075555503368378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,64,51200,0.013906665974193148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,1536,0.006682666639486949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,51200,0.038386666112475924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,128,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,16384,0.015944000747468736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,64,10240,0.007670222057236566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,32,0.005064888960785336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,64,8192,0.007024889190991719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,10240,0.011768889096048145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,64,7168,0.006642666541867786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,65536,0.04792088932461209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,64,6144,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,7168,0.009728888670603434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,64,12288,0.008350222474998897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,64,5120,0.006322666588756774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,6144,0.009015999734401703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,12288,0.013112000293201871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,64,3584,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,4096,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,64,3072,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,64,16384,0.00869422240389718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,3584,0.007284444239404466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,3072,0.0069422221018208405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,64,2048,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,2560,0.006607999818192587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,64,1536,0.004559111264016894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,2048,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,5120,0.008054222497675154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,64,1024,0.003875555677546395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,64,768,0.003526222374704149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,1536,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,64,512,0.0031999999450312722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,1024,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,64,256,0.0028888889484935333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,512,0.00526577772365676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,768,0.008995555341243744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,256,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,64,64,0.0025173332542181015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,128,0.00489688871635331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,64,32,0.002532444480392668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,64,0.0047484445903036325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,8192,0.010384889112578498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,512,0.005283555636803309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,32,51200,0.013799111048380533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,64,4096,0.006310222049554189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,65536,0.048158221774631076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,32,16384,0.0087013335691558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,32,12288,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,51200,0.038712001509136625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,32,10240,0.007698666718271043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,12288,0.013190222283204397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,64,2560,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,10240,0.011752888560295105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,32,7168,0.006659555352396435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,64,128,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,32,6144,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,8192,0.010176888770527309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,32,5120,0.005964444329341252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,7168,0.009725333087974125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,32,4096,0.006640000061856375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,6144,0.009136888715955947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,32,3584,0.006258666515350342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,16384,0.016185777054892648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,5120,0.008086221913496653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,32,3072,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,4096,0.007630222373538547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,32,2560,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,3584,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,32,2048,0.006020444548792309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,2560,0.006654222392373615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,32,1536,0.00423200014564726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,2048,0.006314666734801398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,32,1024,0.0038657776183552216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,1536,0.005962666538026597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,32,768,0.003272888975010978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,1024,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,768,0.0052986666560173035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,768,0.005378666851255629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,32,256,0.0028524444335036804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,512,0.005249777601824866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,32,128,0.0028853333658642242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,256,0.005281777845488654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,32,64,0.0025155554629034465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,128,0.004912000149488449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,32,32,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,64,0.004962666581074397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,3072,0.006984000404675801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,32,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,16384,0.3727235529157851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,51200,0.6113688680860732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,51200,1.1480560302734375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,32,8192,0.007003555695215861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,16384,0.20315644476148817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,12288,0.27941955460442436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,16384,0.13543466726938883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,51200,0.40456355942620176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,12288,0.14916266335381403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,10240,0.20974132749769422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,12288,0.10370666450924343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,8192,0.1733315520816379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,10240,0.08497422271304661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,7168,0.15179377131991917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,32,65536,0.014151111245155334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,6144,0.12984532780117458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,7168,0.09322310818566217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,8192,0.10692622264226277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,7168,0.06154844495985243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,5120,0.10999377568562825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,6144,0.07875288857354058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,10240,0.13249599933624268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,4096,0.09255555603239272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,5120,0.0455351107650333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,8192,0.0715235537952847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,112,32,512,0.0032186667538351486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,32,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,4096,0.0392231113380856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,3072,0.06936444176567926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,3584,0.034322665797339544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,3584,0.04933955603175693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,2560,0.058394663863711886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,3072,0.03090933296415541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,2048,0.047760887278450855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,2560,0.02696177860101064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,5120,0.06673422124650744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,2048,0.023205333285861548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,2048,0.032255109813478254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,3584,0.07980710930294461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,4096,0.05468711256980896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,1024,0.027789334456125896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,1536,0.02533333334657881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,768,0.021348444951905146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,1024,0.015372445185979208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,1024,0.025376889440748427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,512,0.014826666977670459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,2560,0.037061333656311035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,512,0.012631999949614206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,512,0.014130666851997375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,256,0.009343999955389235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,1536,0.01863466699918111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,256,0.011420444481902652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,128,0.008230222596062554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,256,0.01240888900227017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,3072,0.04256888892915514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,128,0.011402666568756104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,768,0.013912889692518445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,32,0.006261333409282897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,6144,0.053411556614769824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,64,0.011400889191362592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,1536,0.03761955433421665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,128,0.011521778172916837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,64,0.005924444645643234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,65536,1.1107003953721788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,768,0.01998044384850396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,16384,0.28032978375752765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,32,0.015478221906556023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,65536,0.41182221306694883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,12288,0.21433422300550672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,65536,0.6280622482299805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,12288,0.12166933218638103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,51200,0.8687697516547309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,10240,0.1951893303129408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,16384,0.10890844133165146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,10240,0.10341778066423203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,8192,0.16046667098999023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,51200,0.47902668846978086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,16384,0.16006755828857422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,8192,0.08539644214842056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,10240,0.06820711162355211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,51200,0.3206053309970432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,6144,0.12045510609944661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,12288,0.08557689189910889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,5120,0.1017964416080051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,6144,0.04972355564435323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,6144,0.0659431086646186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,7168,0.13985599411858454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,4096,0.08275644646750556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,8192,0.05574755536185371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,5120,0.03749155667093065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,7168,0.07221066951751709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,7168,0.057525330119662814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,4096,0.044300443596310086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,3072,0.06392888890372382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,3584,0.02880355715751648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,3584,0.03971199856864081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,2560,0.0532622238000234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,3072,0.025773333178626165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,2048,0.04392533169852363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,2560,0.03030311067899068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,2048,0.019634667370054457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,5120,0.0533884432580736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,2048,0.026583110292752583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,4096,0.03257066673702664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,1536,0.016280000408490498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,1536,0.020993777447276648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,768,0.018858666221300762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,1024,0.013682666752073499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,1024,0.020765332712067496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,512,0.013800889253616333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,768,0.012466666599114736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,2560,0.022893332772784766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,768,0.0165928883685006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,256,0.008809777597586313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,512,0.011400000088744693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,512,0.0124195557501581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,128,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,1536,0.03389511174625821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,3584,0.07306133376227485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,256,0.010634666515721215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,128,0.010159999959998662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,32,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,128,0.010031110710567897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,65536,0.3406702147589789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,3072,0.03495555453830295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,65536,0.2283759911855062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,51200,0.2621013323465983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,256,0.010395555860466426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,64,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,16384,0.08655199739668105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,16384,0.03979466689957513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,16384,0.060673779911465116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,32,0.013228444589508904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,12288,0.06675999694400363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,64,0.009806222385830348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,10240,0.05590933561325073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,12288,0.034210665358437434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,12288,0.045891556474897593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,8192,0.04673066735267639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,10240,0.0274311105410258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,10240,0.03959466682540046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,7168,0.04105066590838962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,8192,0.024301333559883967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,51200,0.17308710681067574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,7168,0.02126577827665541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,7168,0.029682666063308716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,51200,0.11084800296359593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,5120,0.03088088830312093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,6144,0.018901333212852478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,4096,0.024870221813519795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,5120,0.01699199941423204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,5120,0.022936888866954382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,4096,0.015505777464972602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,4096,0.019143111175960965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,65536,0.14241333802541098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,8192,0.03455200129085117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,3072,0.020038222273190815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,3584,0.017278222574128043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,2560,0.016921778519948322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,3072,0.013471110827393003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,3072,0.01585422290696038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,2048,0.013912000589900546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,2560,0.013782222237851886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,3584,0.023027555810080633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,1536,0.011383111278216044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,1024,0.025223111112912495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,6144,0.03550488750139872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,3584,0.013898667362001208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,1024,0.009389333426952362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,1536,0.00943199959066179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,768,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,1024,0.00832799987660514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,6144,0.026418666044871014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,1024,0.010780444575680627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,512,0.005904888941182031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,768,0.008023999631404877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,768,0.008457777400811514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,256,0.004280888785918553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,512,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,2048,0.011047110789351992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,2048,0.012429333395428128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,128,0.003625777860482534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,128,0.006991110742092133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,128,0.0063546668324205614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,1536,0.010622222390439777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,32,0.0035573335157500375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,2560,0.012432888978057437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,64,0.006618666566080517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,65536,0.24591644605000815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,65536,0.17481511169009736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,256,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,64,0.0035457776652442086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,51200,0.18787377410464814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,65536,0.1276737716462877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,16384,0.06408088737063937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,51200,0.13325511084662542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,51200,0.1012844443321228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,12288,0.05013333426581489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,32,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,10240,0.042747557163238525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,12288,0.03649244374699063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,8192,0.03571110963821411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,256,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,10240,0.025389333566029865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,10240,0.0313937763373057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,8192,0.02240000002914005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,8192,0.027689778142505225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,6144,0.027561777167850073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,7168,0.02200355629126231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,7168,0.024051555328898962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,16384,0.03631555702951219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,5120,0.024128889044125874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,6144,0.017869333426157635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,6144,0.02129688858985901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,12288,0.030415111117892798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,5120,0.018413333429230582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,3584,0.017630222770902846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,4096,0.015491555134455362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,7168,0.0313662224345737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,3072,0.015485333071814643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,3584,0.013520888984203339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,3584,0.014261333478821648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,2560,0.013781333135233985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,3072,0.012472000386979846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,16384,0.047967112726635404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,3072,0.013145777914259167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,2048,0.011839110818174152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,512,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,4096,0.020787555310461257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,5120,0.015839111473825242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,2048,0.010852444503042432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,4096,0.014733332726690503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,1536,0.009510222408506606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,1024,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,1024,0.008784888519181145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,768,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,768,0.0076657773719893555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,512,0.005244444641802046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,2560,0.011605333122942181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,1536,0.009703111317422655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,2048,0.010204444328943888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,512,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,512,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,256,0.006711999989218182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,256,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,128,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,64,0.003569777641031477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,768,0.007886221839321984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,32,0.003642666671011183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,64,0.006305777778228124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,256,0.004587555511130227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,1024,0.008251555263996124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,128,0.00388711111413108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,65536,0.16570399867163763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,65536,0.12098133563995361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,128,0.005964444329341252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,16384,0.053745776414871216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,51200,0.12745600276523167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,2560,0.011807111402352652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,51200,0.09647999869452582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,12288,0.04369689027468363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,16384,0.04561777909596761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,1536,0.00868888861603207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,12288,0.028713779317008123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,12288,0.034865776697794594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,8192,0.03096888793839349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,51200,0.1590906646516588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,10240,0.03023911184734768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,7168,0.02697244452105628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,8192,0.02141422198878394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,8192,0.02647200061215295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,6144,0.02409244411521488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,7168,0.020781333247820537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,7168,0.022679999470710754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,6144,0.017318222257826064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,5120,0.0207368897067176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,32,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,10240,0.023952888117896184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,5120,0.01752622259987725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,4096,0.015207999282413058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,4096,0.013878222141000958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,3584,0.015146666102939181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,65536,0.20674310790167916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,5120,0.01605422298113505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,3584,0.014070222775141398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,3072,0.013452444639470843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,3072,0.012124444047609964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,2560,0.012129777835475074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,3072,0.012847999731699625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,2560,0.010599111517270407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,2048,0.01070844464831882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,10240,0.03617066807217068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,2048,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,3584,0.013196444345845116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,4096,0.017521777086787753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,1024,0.007584888901975419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,1536,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,1536,0.00903555585278405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,768,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,1024,0.008405333591832055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,2560,0.011737777955002256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,16384,0.03581510980923971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,768,0.007231111327807109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,512,0.007055111229419708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,256,0.003924444317817688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,256,0.005991999887757831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,1024,0.007713778151406183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,128,0.003896000070704354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,256,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,768,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,6144,0.01999288962946998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,64,0.003212444484233856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,32,0.00360355567600992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,2048,0.01072533345884747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,512,0.006268444574541516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,64,0.006063111126422882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,65536,0.17628978358374703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,51200,0.1417493290371365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,65536,0.07120444377263387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,512,0.00490488898422983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,128,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,1536,0.009156444834338294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,32,0.006646222124497096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,12288,0.03939111034075419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,65536,0.12240621778700088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,12288,0.027154665854242113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,12288,0.018610666195551556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,10240,0.03296711047490438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,51200,0.09553688764572144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,16384,0.049478222926457725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,10240,0.016527111331621807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,10240,0.02335111134582096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,7168,0.02472355630662706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,8192,0.014820444915029736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,7168,0.01351999988158544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,6144,0.021577777134047613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,128,0.006837333242098491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,6144,0.01608800060219235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,5120,0.018792000081804063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,16384,0.03439022103945414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,5120,0.014135999812020196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,5120,0.011744000017642975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,7168,0.017581333716710407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,4096,0.01072088877360026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,4096,0.01240533341964086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,51200,0.05777511331770155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,8192,0.028160889943440754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,3072,0.012582222620646158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,6144,0.012799999780125089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,3584,0.009758222434255812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,2560,0.011108444796668159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,3072,0.008803555534945594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,8192,0.01998666591114468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,3072,0.010886222124099731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,2560,0.00830488900343577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,4096,0.016177778442700703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,2560,0.010035555395815108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,2048,0.007696888513035244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,2048,0.00907555553648207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,1024,0.006666666517655055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,1536,0.007335999773608313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,1536,0.007318221860461765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,768,0.0052800000541739995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,1024,0.00665511108107037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,1024,0.007044444481531779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,512,0.0046364445653226645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,16384,0.02308711078431871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,768,0.006611555400821898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,2048,0.009903999666372934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,512,0.006309333360857434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,1536,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,128,0.0032951111594835916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,256,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,3584,0.014116444521480136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,128,0.005630222045713001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,64,0.0031999999450312722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,3584,0.011727111207114326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,768,0.00665155549844106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,256,0.003922666526503033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,64,0.005888888819350137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,512,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,256,0.005979555762476391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,65536,0.11969155735439724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,128,0.006004444426960415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,51200,0.05593066745334201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,32,0.003547555456558863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,16384,0.04773599902788798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,16384,0.02233422299226125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,65536,0.16044444508022732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,12288,0.03754933343993293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,51200,0.12740888860490587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,65536,0.06866044468349881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,10240,0.03160444564289517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,12288,0.02604088849491543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,8192,0.027140445179409448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,51200,0.0923902193705241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,10240,0.022555554906527202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,7168,0.023778667052586872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,8192,0.014256000518798828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,7168,0.013391111459996967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,6144,0.02088000045882331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,7168,0.01720888912677765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,12288,0.018029333816634286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,6144,0.012466666599114736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,6144,0.015783111254374187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,4096,0.015101333459218344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,5120,0.011412444214026133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,10240,0.016022221909628976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,5120,0.013959111438857185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,3584,0.012764444781674279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,4096,0.010040000081062317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,4096,0.012376888758606382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,3584,0.00904977818330129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,3584,0.011737777955002256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,32,0.006313777632183499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,2560,0.01036622209681405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,3072,0.008117333054542542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,2048,0.009379555781682333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,2560,0.008008889026112026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,2560,0.009747555686367882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,1536,0.007829333345095316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,2048,0.008325333396593729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,1536,0.0070062221752272705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,1024,0.0064533332155810455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,3072,0.011398221883508893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,16384,0.033591111501057945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,8192,0.019465777609083388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,768,0.0052355556852287715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,1024,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,3072,0.01071111112833023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,512,0.004579555657174853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,5120,0.017647110753589206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,256,0.0038986665507157645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,512,0.0063075555695427795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,2048,0.007640889121426477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,512,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,256,0.005976888868543837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,256,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,64,0.003185777821474605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,1536,0.007346666521496243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,128,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,1024,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,64,0.005704000178310607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,32,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,65536,0.1442533334096273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,768,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,51200,0.11454399426778157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,65536,0.0663013325797187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,16384,0.043807999955283276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,51200,0.08768800232145522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,51200,0.054099556472566396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,128,0.005980444451173146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,12288,0.036432888772752546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,32,0.0032515554792351196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,10240,0.0342497792508867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,12288,0.017464889420403373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,12288,0.02496888902452257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,8192,0.021878222624460857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,10240,0.02160266704029507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,768,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,7168,0.018619555566045973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,8192,0.013827555709415011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,7168,0.012897777888509961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,7168,0.016908443636364408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,16384,0.03270399901602004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,6144,0.012079111403889127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,6144,0.015307555596033732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,5120,0.014467555615637036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,5120,0.010520888699425591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,65536,0.11461777157253689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,4096,0.013357333011097379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,5120,0.013570666313171387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,3584,0.01144088887506061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,4096,0.009344889058007134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,4096,0.012087999946541257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,3584,0.011433777709801992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,3072,0.010379555324713389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,10240,0.015680889288584392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,3072,0.008294222255547842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,3072,0.010593777729405297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,128,0.003576000117593341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,8192,0.019016888406541612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,2048,0.008630221916569604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,2560,0.007974222302436829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,2560,0.009380444056457942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,16384,0.02141511109140184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,1536,0.006949333681000604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,2048,0.008053333395057255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,1024,0.005590222362014983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,1536,0.0069546666410234236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,1536,0.006990222467316522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,768,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,1024,0.006653333289755716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,1024,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,512,0.004522666749027041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,2560,0.009343999955389235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,768,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,256,0.003587555554178026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,512,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,6144,0.016947555873129103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,2048,0.007536888950400882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,128,0.0032479998966058097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,256,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,256,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,3584,0.00867377801073922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,32,0.0031973332580592898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,128,0.005290666802061929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,64,0.0057813334796163775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,65536,0.11527910497453477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,32,0.006059555543793573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,51200,0.09275377458996242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,65536,0.06361066632800631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,16384,0.03638755612903171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,51200,0.08482666810353597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,64,0.0032222221295038858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,16384,0.020398222737842135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,128,0.005990222096443176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,12288,0.01711466742886437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,12288,0.0240346673462126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,768,0.006342222293217977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,512,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,10240,0.01516177753607432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,65536,0.10930399762259589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,10240,0.020840888222058613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,51200,0.051853332254621715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,8192,0.01346577786737018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,8192,0.018745778335465323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,16384,0.0310808883772956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,6144,0.014518222875065274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,7168,0.016516443755891588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,5120,0.012868444124857584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,6144,0.011447110937701331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,6144,0.014808000789748298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,10240,0.02684088879161411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,4096,0.011615110768212212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,8192,0.022666666242811415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,4096,0.008714666797055138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,3584,0.01072088877360026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,4096,0.011880888707107969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,3072,0.009876444107956356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,3584,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,3584,0.011160000330872007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,2560,0.009355555805895064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,3072,0.009696889254781935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,2560,0.007698666718271043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,2560,0.008631111019187504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,5120,0.013464888764752282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,2048,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,1536,0.006611555400821898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,12288,0.030348443322711523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,3072,0.008105778031879002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,1024,0.005584888988071018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,1536,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,1024,0.006593777901596493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,2048,0.00833422193924586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,1024,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,7168,0.016546666622161865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,512,0.00423200014564726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,7168,0.012418666647540199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,1536,0.007037333316273159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,256,0.0036888888312710654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,512,0.006126222097211414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,768,0.004906666775544484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,256,0.005492444253630108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,768,0.0063475556671619415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,2048,0.007825777762466008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,64,0.0028728888266616394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,128,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,32,0.0032417778339650896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,512,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,64,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,65536,0.09308889177110459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,32,0.005974222388532426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,5120,0.010040889183680216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,65536,0.06174221966001722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,65536,0.0679324467976888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,768,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,128,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,16384,0.03016800019476149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,51200,0.0529093345006307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,12288,0.025808889004919264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,16384,0.019959110352728102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,16384,0.02033866610791948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,10240,0.023039999935362074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,12288,0.016407999727461074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,51200,0.07561155822541979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,128,0.0032195556494924757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,10240,0.014753777119848462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,10240,0.014534221755133735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,7168,0.014460444450378418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,8192,0.013046222428480784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,8192,0.013572444518407186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,6144,0.012880889077981314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,7168,0.012076444096035428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,7168,0.011752888560295105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,5120,0.011384000380833944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,6144,0.010712888505723743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,6144,0.011077333655622272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,12288,0.016427555018001132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,5120,0.009335111412737105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,5120,0.010136889086829292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,8192,0.015949333707491558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,3584,0.009339555270142024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,4096,0.008838222258620793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,3072,0.008464888566070134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,51200,0.04962044292026096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,2560,0.007710222568776872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,3072,0.00795022232664956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,3584,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,2048,0.006736889067623351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,2560,0.0074879998962084455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,3072,0.007353777686754863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,2048,0.007318221860461765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,2048,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,4096,0.010988444089889526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,1536,0.006984000404675801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,1024,0.004567999806669024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,1536,0.006251555350091722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,768,0.0042239998777707415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,1024,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,1024,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,512,0.0035955554081334006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,3584,0.00811822215716044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,768,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,256,0.003552000141806073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,512,0.006004444426960415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,512,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,128,0.00318755561278926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,256,0.005920889063013925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,256,0.00509688879052798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,128,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,256,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,128,0.004967111266321606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,32,0.0029084444459941653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,64,0.0053031109273433685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,32,0.005608888963858287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,65536,0.08830666542053223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,51200,0.07106133302052815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,1536,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,65536,0.06536799669265747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,16384,0.029509332444932725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,2560,0.007086222370465596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,51200,0.05113510953055488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,64,0.00319377767542998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,12288,0.023769777682092454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,4096,0.008685333033402761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,16384,0.0199946661790212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,12288,0.01612355477280087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,768,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,8192,0.01584533353646596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,10240,0.014473777678277759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,10240,0.014141332772043018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,7168,0.013790222505728403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,8192,0.012530666258600024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,65536,0.06083110968271891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,8192,0.012772444221708508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,6144,0.012533333566453723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,7168,0.011425777441925473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,5120,0.01106844428512785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,7168,0.011519999967681037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,6144,0.010713777608341642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,4096,0.010075555907355415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,16384,0.01957244508796268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,5120,0.009365333451165093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,10240,0.021249777740902368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,3584,0.008999110923873054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,5120,0.009914666414260864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,12288,0.01575911045074463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,3072,0.008110221889283922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,4096,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,2560,0.007463999920421177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,3584,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,3072,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,2048,0.006320000108745363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,2560,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,2560,0.0069057775868309875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,51200,0.04880355464087593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,2048,0.007295110987292395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,2048,0.006454222318198945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,6144,0.010767999622556897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,1024,0.004559999952713649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,1536,0.006983111302057902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,4096,0.008564444051848518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,1536,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,768,0.003926222109132343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,1024,0.006299555715587404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,1024,0.005639111002286275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,768,0.006324444380071428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,3072,0.007340444458855524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,768,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,256,0.0035457776652442086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,512,0.006014222072230444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,512,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,256,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,1536,0.0052062223354975385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,64,0.002990222225586573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,128,0.004942222187916438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,32,0.0029022221763928733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,3584,0.00833777752187517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,64,0.0053004444473319584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,512,0.003544888976547453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,32,0.0052897776994440295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,51200,0.06281155347824097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,65536,0.05957688887914022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,128,0.003228444399105178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,16384,0.024719999896155462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,256,0.004952888935804367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,51200,0.04927466644181145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,12288,0.02093511157565647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,16384,0.01922044489118788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,10240,0.018349332941903006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,12288,0.015285332997639975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,10240,0.014319111903508505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,10240,0.013732444081041547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,65536,0.06262666649288602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,51200,0.04764444298214383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,8192,0.012016889121797351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,7168,0.014446222119861178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,16384,0.019304000669055514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,6144,0.013464888764752282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,7168,0.011333333121405708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,65536,0.07742310894860162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,5120,0.01071733319097095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,6144,0.01053422192732493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,8192,0.016553777787420485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,4096,0.009355555805895064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,128,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,4096,0.007956444389290279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,3584,0.008672888908121321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,7168,0.010767999622556897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,12288,0.015858666764365304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,3584,0.008336000144481659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,3584,0.007650666766696506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,6144,0.010042666561073726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,2560,0.008020444048775567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,3072,0.0076657773719893555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,5120,0.009376889301670922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,2048,0.0069368887278768755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,2560,0.007672000262472365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,2560,0.0069288888739215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,8192,0.012553777959611682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,1536,0.005631111148330901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,2048,0.006304889089531369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,4096,0.008402666283978356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,1536,0.006055111272467508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,3072,0.00775999989774492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,768,0.004582222137186262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,1024,0.0063457778758472866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,1024,0.00562755556570159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,512,0.004226666771703296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,768,0.006324444380071428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,768,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,256,0.0035537779331207275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,512,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,2048,0.00720266666677263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,128,0.003228444399105178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,256,0.005888888819350137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,5120,0.009582222335868413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,256,0.005055111315515307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,64,0.0028764444092909494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,32,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,128,0.004967111266321606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,65536,0.06632355848948161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,32,0.005283555636803309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,512,0.00599644457300504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,65536,0.060864885648091636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,51200,0.05454044540723165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,1024,0.005120000077618493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,51200,0.04604622059398227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,128,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,64,0.004943999979231092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,16384,0.018582221534517076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,3072,0.007048889166778988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,12288,0.015847111741701763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,65536,0.05817066960864597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,12288,0.01513155632548862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,10240,0.016809778081046212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,1536,0.00701333334048589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,51200,0.04734933376312256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,8192,0.014782221780882942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,16384,0.01908533275127411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,8192,0.012084444363911947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,7168,0.011038222246699862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,7168,0.010728889041476779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,10240,0.013528000149461957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,6144,0.012732444538010491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,10240,0.013482666677898832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,5120,0.011566222541862063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,8192,0.01146133326821857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,12288,0.019347555107540555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,4096,0.009538667069541084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,5120,0.009386666946940953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,5120,0.009073778159088558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,3584,0.008364444805516137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,4096,0.008354666332403818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,4096,0.008016889293988546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,3072,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,3584,0.008224888808197444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,3584,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,16384,0.022220444348123338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,3072,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,6144,0.010051555931568146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,2560,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,7168,0.013924444715181986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,2560,0.006680000159475539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,1536,0.005295111073387994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,2048,0.007122666471534305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,2048,0.006364444477690591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,1024,0.004557333472702238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,1536,0.006813333266311222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,768,0.004002666721741359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,1024,0.006393777827421824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,1024,0.005649777750174205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,2560,0.00665155549844106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,3072,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,6144,0.01032266683048672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,256,0.003569777641031477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,256,0.0052844443255000645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,256,0.005727111051479976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,128,0.0032151111712058387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,512,0.003705777641799715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,2048,0.005569777968857024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,128,0.005895111295912001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,128,0.004948444250557157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,512,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,64,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,65536,0.05486399928728739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,1536,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,65536,0.056279109583960645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,768,0.005611555443869696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,51200,0.04449333416091072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,64,0.0028862222615215513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,51200,0.04543911086188423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,512,0.0053280000057485365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,51200,0.04535911149448819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,12288,0.015570667054918079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,16384,0.018619555566045973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,16384,0.018186666899257235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,32,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,12288,0.014856000741322836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,12288,0.014760888285107084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,8192,0.01237511138121287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,10240,0.012663111090660095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,768,0.006264888991912206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,7168,0.012196444802814059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,8192,0.011298666397730509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,8192,0.011445333560307821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,6144,0.01202222208182017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,7168,0.010442666709423065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,7168,0.010623999767833285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,6144,0.009727111293209923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,6144,0.009344889058007134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,10240,0.013784888717863293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,4096,0.00889155517021815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,65536,0.057929780748155385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,4096,0.007689778175618913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,4096,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,3584,0.00776711106300354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,32,0.003240888938307762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,3072,0.007060444189442529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,3584,0.00830488900343577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,5120,0.010016000105275048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,2560,0.006266666783226862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,3072,0.007711999946170383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,3072,0.007048000064161088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,2048,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,2560,0.007686222592989604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,2560,0.006781333436568578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,10240,0.013120888835854001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,2048,0.007134222322040134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,2048,0.006341333190600078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,1024,0.004229333251714706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,1536,0.00693866651919153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,1024,0.006404444575309753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,3584,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,1024,0.005623111294375525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,5120,0.009033777647548253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,768,0.005620444400442972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,1536,0.004875555634498596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,512,0.0034977777136696708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,512,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,256,0.0032791110376516977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,768,0.00387999994887246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,256,0.005229333208666907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,128,0.0028782222006056043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,5120,0.0087333329849773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,768,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,64,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,16384,0.018960000740157235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,1536,0.005971555494599872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,64,0.004935111022657818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,65536,0.042374223470687866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,32,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,256,0.006154666758245892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,51200,0.03429333368937174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,128,0.005809777726729711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,128,0.004984000076850255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,51200,0.04491110973887973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,32,0.0028844444702068963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,12288,0.013758222262064615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,12288,0.014202665951516895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,65536,0.05543022354443868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,10240,0.013200889031092325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,12288,0.013825777504179211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,16384,0.015377778146002026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,10240,0.012472000386979846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,51200,0.04404533240530226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,10240,0.012453333371215396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,7168,0.011752000285519494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,8192,0.010749333434634738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,7168,0.010430222584141625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,6144,0.011018666956159802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,65536,0.055496888028250806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,6144,0.009711999860074785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,8192,0.011742221812407175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,512,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,16384,0.017609778377744887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,4096,0.008683555656009251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,5120,0.008567111359702216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,7168,0.010036444498433007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,3584,0.007082666787836287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,3584,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,5120,0.009926222264766693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,3072,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,3584,0.008013333711359236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,16384,0.017695999807781644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,2560,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,3072,0.007654222349325816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,3072,0.0069875551594628235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,2048,0.0052373334765434265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,2560,0.007328888608349695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,4096,0.007684444387753804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,1536,0.004947555561860402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,2048,0.007144889069928064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,2048,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,1024,0.004233777936961916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,6144,0.009093333449628618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,1536,0.005985777825117111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,768,0.0038782221575578055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,1024,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,1024,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,512,0.003551111039188173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,5120,0.009010666774378883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,768,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,256,0.0032293332947625052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,4096,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,2560,0.006667555620272954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,128,0.0030737777964936364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,256,0.005021333280536863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,1536,0.006711999989218182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,64,0.002891555635465516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,128,0.005991111199061076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,128,0.005079111291302575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,32,0.0028880000528362063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,8192,0.01107377807299296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,65536,0.03130577670203315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,64,0.004935111022657818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,768,0.006331555545330048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,32,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,65536,0.0548622210820516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,512,0.005984000033802456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,65536,0.05275466707017687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,512,0.0052897776994440295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,12288,0.013955555028385587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,51200,0.04216000106599596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,16384,0.016185777054892648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,10240,0.012528888881206512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,16384,0.017082666357358296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,12288,0.01365244471364551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,8192,0.011391110718250275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,10240,0.012256888879670037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,10240,0.011809777882364062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,7168,0.010203555226325989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,51200,0.025197333759731714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,16384,0.01314755529165268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,7168,0.009699555734793345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,6144,0.010586666564146677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,7168,0.010398222340477837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,6144,0.009732444253232744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,5120,0.009472000102202097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,8192,0.011272889044549731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,8192,0.010401777923107147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,4096,0.007968000239796108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,5120,0.009000889129108852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,51200,0.04416799876425001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,4096,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,256,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,4096,0.007682667010360294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,3072,0.006599999964237213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,3584,0.008036444584528605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,3584,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,2560,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,3072,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,3072,0.0070168889231152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,2048,0.006321777900060018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,5120,0.008562666674455007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,2560,0.006702222343948152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,1536,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,2048,0.007053333024183909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,3584,0.007300444775157505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,1024,0.004243555582231945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,1536,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,6144,0.009120889008045197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,1024,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,1024,0.005635555419656966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,768,0.006242666807439592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,2560,0.007387555307812161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,768,0.0052800000541739995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,256,0.0032097777972618737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,512,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,512,0.00526577772365676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,128,0.002899555489420891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,256,0.0052782222628593445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,128,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,768,0.0042364444169733255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,512,0.003565333369705412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,32,0.0028853333658642242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,64,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,768,65536,0.024694222542974684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,2048,0.0063279999627007385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,32,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,768,51200,0.02199199961291419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,1536,0.006783111227883234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,65536,0.05473866727617052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,65536,0.0515342222319709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,768,16384,0.01349866638580958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,128,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,51200,0.04179199867778354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,16384,0.016189333465364244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,12288,0.01349688900841607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,768,10240,0.011427555647161273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,256,0.005892444401979446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,10240,0.012431110772821637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,768,8192,0.009345778160625035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,12288,0.013189333180586496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,768,12288,0.011885333392355176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,768,7168,0.00904177791542477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,8192,0.011076444553004371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,8192,0.010428444378905825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,768,6144,0.008370666868156856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,7168,0.010339555641015371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,7168,0.009691555466916826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,768,5120,0.008017777568764156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,6144,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,6144,0.009039999710188972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,768,4096,0.007763555480374231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,5120,0.009000889129108852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,5120,0.008381333616044786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,768,3584,0.007663110891977946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,4096,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,16384,0.016208888755904306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,768,3072,0.0063528890411059065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,3584,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,3584,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,768,2560,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,3072,0.0070186663005087115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,2560,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,2560,0.006610666712125142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,10240,0.011796444654464722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,2048,0.0069982219073507525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,4096,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,768,1536,0.00462755560874939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,1536,0.006671111202902264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,3072,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,1024,0.00563466673096021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,1024,0.006351111249791251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,768,768,0.00360000009338061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,2048,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,768,0.006311111152172089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,768,512,0.003253333270549774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,768,0.005574222240183089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,512,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,768,256,0.0029226665695508323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,768,2048,0.0052791109515561005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,256,0.004979555391603046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,256,0.005660444498062134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,768,128,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,768,1024,0.0038728887836138406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,128,0.005629333357016246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,128,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,768,64,0.002879111096262932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,768,32,0.0028728888266616394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,64,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,51200,0.04418399930000305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,32,0.004958222309748332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,65536,0.053364445765813194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,512,51200,0.016694222887357075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,512,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,51200,0.04090933336151971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,512,16384,0.011297778122954898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,51200,0.03938311007287767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,64,0.0028862222615215513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,16384,0.016186666157510545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,512,12288,0.009771555662155151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,16384,0.015853333804342482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,512,10240,0.009694221946928237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,12288,0.013432888521088494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,65536,0.05123200019200643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,512,8192,0.008386666576067606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,10240,0.011658666862381829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,512,7168,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,8192,0.01109866632355584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,8192,0.010449777874681683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,512,6144,0.010042666561073726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,512,65536,0.02029244436158074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,7168,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,6144,0.009287111461162567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,12288,0.013456888496875763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,10240,0.012415111064910889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,512,4096,0.007698666718271043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,5120,0.008399110701349046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,4096,0.008416000339719985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,7168,0.01035288886891471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,512,3584,0.0069857777820693115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,3584,0.008023999631404877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,512,3072,0.006581333362393909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,1536,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,3072,0.0070088886552386815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,512,2560,0.006161777923504512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,4096,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,2560,0.0068604445291890045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,512,2048,0.005232000102599462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,2560,0.007636444436179266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,6144,0.009687110781669617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,2048,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,12288,0.013088000317414602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,5120,0.00904088881280687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,1536,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,512,1024,0.0038977778620190094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,1024,0.006304889089531369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,512,768,0.003544888976547453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,3584,0.007359999749395583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,512,1536,0.0047253333032131195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,512,512,0.0035084444615576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,2048,0.006314666734801398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,3072,0.007666666474607255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,512,256,0.0032115555885765287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,512,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,256,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,1024,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,256,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,128,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,512,5120,0.008671999805503422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,128,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,512,32,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,768,0.006319111006127463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,256,65536,0.016185777054892648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,64,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,32,0.004975999808973736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,256,51200,0.018576888574494254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,65536,0.05014666583802965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,512,128,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,256,16384,0.009739555418491364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,51200,0.03785955574777391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,51200,0.039566222164365984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,256,12288,0.009752000371615091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,16384,0.016167999969588388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,16384,0.015501333607567681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,256,10240,0.009090666969617208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,12288,0.012835555606418185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,10240,0.0124195557501581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,65536,0.04723199870851305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,256,8192,0.00850044439236323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,8192,0.011090666883521609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,256,7168,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,8192,0.01016622202263938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,512,64,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,768,0.005618666609128316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,7168,0.01034400032626258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,7168,0.009692444569534725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,12288,0.013502221968438892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,256,5120,0.008715555899673039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,1536,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,256,4096,0.007630222373538547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,5120,0.009019555317031013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,10240,0.011523555550310345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,5120,0.008359111017651027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,256,3584,0.006732444382376141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,4096,0.008325333396593729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,4096,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,256,3072,0.006259555617968242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,3584,0.007991111112965478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,3584,0.007359999749395583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,256,2560,0.005629333357016246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,3072,0.007671111159854465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,3072,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,6144,0.009705777797434065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,512,0.005994666781690385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,2560,0.006647999915811751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,256,1536,0.004295111116435793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,2048,0.007044444481531779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,2048,0.006260444306664997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,256,1024,0.0038737778862317405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,1536,0.006664000037643645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,1536,0.005971555494599872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,1024,0.006509333435032103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,256,6144,0.0086666668454806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,768,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,768,0.005628444254398346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,256,512,0.0032391111469931076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,6144,0.009045333498054082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,256,256,0.0028693332440323303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,512,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,512,0.005327111317051782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,256,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,256,0.005127999931573868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,256,64,0.0028391112056043413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,256,768,0.003583999971548716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,128,0.005965333431959152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,1024,0.00534400012758043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,128,65536,0.014736889137162102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,64,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,256,2048,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,128,51200,0.013374221821626028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,65536,0.04656533400217692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,51200,0.03772355450524224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,51200,0.0380586650636461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,256,32,0.002581333327624533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,16384,0.01571644511487749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,128,12288,0.009699555734793345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,16384,0.016184889607959323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,2560,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,128,10240,0.009012444151772393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,12288,0.01349866638580958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,65536,0.04724533359209696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,12288,0.012780444489585029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,128,8192,0.008000888758235509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,10240,0.012136889000733694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,256,128,0.0029173334025674393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,10240,0.011429333024554782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,8192,0.010795555180973478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,128,6144,0.008691555923885768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,7168,0.010104888843165504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,128,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,128,5120,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,6144,0.00868800034125646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,32,0.004971555537647671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,128,4096,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,5120,0.008991111483838823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,4096,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,4096,0.007667555577225155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,8192,0.010404444403118556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,7168,0.009711999860074785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,3584,0.008006222546100616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,128,3072,0.006305777778228124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,3072,0.007689778175618913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,3072,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,5120,0.008352888955010308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,2560,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,128,7168,0.00904622260067198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,3584,0.007324444750944774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,6144,0.009709333380063375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,128,1536,0.004564444637960858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,2048,0.00703111125363244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,128,2560,0.005620444400442972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,1536,0.006715555571847492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,1536,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,1024,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,2560,0.006660444455014334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,128,768,0.0034933334423436057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,1024,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,128,512,0.0032346666687064697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,768,0.006107555495368109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,2048,0.0063279999627007385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,128,256,0.002895999906791581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,512,0.0052675555149714155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,256,0.005687110953860813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,128,1024,0.003879111260175705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,128,16384,0.009345778160625035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,128,3584,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,128,0.005127111242877113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,128,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,128,32,0.002522666628162066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,768,0.005591111050711737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,64,0.005081777771313985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,64,65536,0.01479022204875946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,32,0.004909333255555895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,64,51200,0.014139556222491793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,256,0.004983110974232356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,64,16384,0.009551111194822524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,65536,0.0473599996831682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,64,12288,0.009689778089523315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,128,2048,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,128,64,0.002518222149875429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,16384,0.015831111205948722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,64,8192,0.008029333419269985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,512,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,64,7168,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,8192,0.010427555276287926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,64,6144,0.007346666521496243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,7168,0.009482666850090027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,64,5120,0.006727111008432176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,51200,0.03808622227774726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,64,10240,0.009024000002278222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,12288,0.012818666795889536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,64,3584,0.0069751110341813825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,10240,0.011415999796655444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,64,3072,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,3584,0.007380444142553542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,64,2560,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,64,2048,0.005978666659858491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,6144,0.009038222332795462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,2560,0.006671999891599019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,5120,0.008341333104504479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,64,1024,0.003565333369705412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,4096,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,64,768,0.003547555456558863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,1024,0.005617777920431561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,64,512,0.003206222214632564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,3072,0.006811555474996567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,64,256,0.0028515555378463534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,64,4096,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,64,1536,0.004604444321658877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,256,0.004934222333961063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,2048,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,1536,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,64,32,0.0028986665937635633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,64,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,128,128,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,768,0.00546488869521353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,32,51200,0.019463999403847587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,65536,0.0469413333468967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,512,0.00509066672788726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,32,16384,0.009352000223265754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,32,12288,0.009696000152164036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,51200,0.038140445947647095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,64,64,0.002551111082235972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,32,10240,0.008656000097592672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,32,8192,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,128,0.005025777965784073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,32,7168,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,10240,0.011431111229790581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,32,65536,0.022375111778577168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,32,0.004979555391603046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,8192,0.010382222632567087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,6144,0.00904177791542477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,64,128,0.002862222285734283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,32,4096,0.007040888898902469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,32,3584,0.006612444503439798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,5120,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,32,3072,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,16384,0.015483554866578845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,32,2560,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,12288,0.012785777449607849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,3584,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,32,2048,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,32,1536,0.004300444490379757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,2560,0.006524444454246097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,32,6144,0.006990222467316522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,7168,0.009733333355850643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,32,768,0.0035662220584021676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,1024,0.005286222116814719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,32,512,0.003216000066863166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,4096,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,32,256,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,512,0.005063999858167437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,32,128,0.0025724443710512584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,32,1024,0.0038844446341196695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,32,5120,0.00666844430896971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,32,64,0.0025173332542181015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,1536,0.005964444329341252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,97,32,32,0.002616000051299731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,128,0.004612444589535396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,64,0.004986666556861666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,768,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,3072,0.006991110742092133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,51200,0.6081342167324489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,51200,1.1509875191582573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,16384,0.37499909930759007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,16384,0.20224977864159477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,51200,0.3985866705576579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,2048,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,12288,0.15678844186994764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,32,0.0048942222363419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,12288,0.10183466805352105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,16384,0.13918044832017687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,256,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,16384,0.2907617886861165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,10240,0.13088177310095891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,12288,0.2213724454243978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,8192,0.17496266629960802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,12288,0.28249510129292804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,8192,0.107479108704461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,10240,0.20875910917917886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,10240,0.19144622484842935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,7168,0.09042133225335015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,8192,0.07482310798433092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,8192,0.14789777331882054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,6144,0.12933066156175402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,7168,0.062126219272613525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,7168,0.13599467277526855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,6144,0.05875200033187866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,7168,0.15088889333936903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,5120,0.06677244769202338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,10240,0.08518133560816447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,4096,0.09434044361114502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,5120,0.0980257789293925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,4096,0.05469155642721388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,6144,0.07885689205593534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,4096,0.041834665669335254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,5120,0.10956711239284939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,4096,0.07893333170149061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,3072,0.0694133308198717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,3584,0.04866755670971341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,3584,0.037096887826919556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,6144,0.11633333894941543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,51200,0.8901644812689887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,2560,0.05836977561314901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,3072,0.031713777118259005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,3072,0.06248888704511854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,2560,0.028985778490702312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,2048,0.04746311240726047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,2560,0.053977777560551964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,2048,0.03205510973930359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,1536,0.03743822375933329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,2048,0.04634222057130602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,3584,0.07074311044481066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,3072,0.04244266615973579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,1536,0.019426667028003268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,1024,0.02882222334543864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,2560,0.0370142228073544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,3584,0.07976533306969537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,1024,0.015319110618697273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,1024,0.032067557175954185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,2048,0.0237128883600235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,5120,0.04705244302749634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,512,0.014529777897728814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,768,0.02758311066362593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,1536,0.02511377798186408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,1536,0.038755555947621666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,512,0.01276000009642707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,512,0.02334755493534936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,1024,0.02664799988269806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,256,0.012429333395428128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,128,0.00795555528667238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,768,0.020452444752057392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,128,0.01143200033240848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,128,0.016363556186358135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,128,0.011485333244005838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,64,0.0069884442620807225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,32,0.007012444237867991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,768,0.02129688858985901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,64,0.011403555671374003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,256,0.011139555937714048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,32,0.015547555353906421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,256,0.018930666976504855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,512,0.015768888923856948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,256,0.0092275556590822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,65536,1.0928604337904189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,51200,0.860031975640191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,768,0.014182221558358936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,51200,0.47655291027492946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,65536,0.4473119841681586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,16384,0.16007999579111734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,65536,0.6308986875745985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,51200,0.334408892525567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,16384,0.11145688427819146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,12288,0.21045866277482775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,12288,0.12546400229136148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,10240,0.17046577400631377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,12288,0.08628888924916585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,16384,0.2776675489213732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,8192,0.14099289311303034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,10240,0.06978755527072482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,10240,0.10063111119800144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,16384,0.23863021532694498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,7168,0.12133333418104385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,12288,0.17863821983337402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,8192,0.08534044689602321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,8192,0.12130578358968098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,7168,0.07242933246824476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,6144,0.1050986647605896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,7168,0.10689600308736165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,6144,0.06331555710898505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,6144,0.09321066406038071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,10240,0.1536853313446045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,7168,0.05145066645410326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,5120,0.05340355634689331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,51200,0.7341813511318631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,4096,0.07621777719921537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,5120,0.03820888863669501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,5120,0.08063111040327284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,65536,0.9806195365058051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,4096,0.04651555418968201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,4096,0.03284444411595663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,3584,0.06405866808361478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,4096,0.06709244516160753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,3072,0.05515999926461113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,3584,0.03962577713860406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,3584,0.061271111170450844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,2560,0.047134220600128174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,6144,0.04503733250829908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,3072,0.03480089041921828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,3072,0.026199999782774184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,2048,0.03864622116088867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,2560,0.023082666926913794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,5120,0.08998844358656143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,2048,0.026583110292752583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,2048,0.019705777366956074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,3584,0.03199822372860379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,2048,0.04020622372627258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,1536,0.021020443903075323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,1024,0.023207111491097346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,1536,0.03384266628159417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,2560,0.030419554975297716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,1024,0.020830222302012973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,768,0.01686044368478987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,1024,0.027607111467255488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,1536,0.03052266769938999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,2560,0.047078222036361694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,512,0.012068444656001197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,768,0.01720355616675483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,1536,0.01663466625743442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,512,0.013271110753218332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,512,0.011168888873524137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,256,0.008626666333940294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,3072,0.05424799852901035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,1024,0.013749333719412485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,128,0.005583999885453119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,256,0.017129777206314933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,768,0.012506666282812754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,128,0.010002666877375709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,128,0.014710222681363424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,64,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,32,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,512,0.020383111304706998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,64,0.009760888914267222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,32,0.013117333253224691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,65536,0.34093332290649414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,256,0.010109333528412713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,65536,0.2266622119479709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,51200,0.2690391010708279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,768,0.023932443724738225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,65536,0.14685332775115967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,51200,0.1735617717107137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,16384,0.09082311391830444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,51200,0.11642666657765706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,256,0.010379555324713389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,16384,0.06059110826916165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,16384,0.04252533449067009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,12288,0.07098933060963948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,16384,0.058449778291914195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,8192,0.0614853302637736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,12288,0.03491111265288459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,65536,0.19697866174909803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,12288,0.043771555026372276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,51200,0.15402666727701822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,8192,0.05029066734843784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,10240,0.02868710954984029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,10240,0.03822044531504313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,7168,0.04236711064974467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,8192,0.025795555777019922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,8192,0.033956444925732084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,128,0.010301333334710863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,7168,0.029940442906485662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,6144,0.03677599959903293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,10240,0.03963111175431146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,6144,0.026612444056404963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,5120,0.03187822302182516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,6144,0.01975644462638431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,8192,0.033915556139416166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,5120,0.018734221657117207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,5120,0.02296177711751726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,4096,0.02685600022474925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,5120,0.02334044377009074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,7168,0.022125333547592163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,3584,0.02344088918632931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,4096,0.019190222024917603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,4096,0.020456888609462313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,3584,0.01725955473052131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,3072,0.021381333470344543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,3584,0.018741332822375827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,3072,0.015543111496501498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,3072,0.017261332935757108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,4096,0.01570399933391147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,2560,0.017813333206706576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,6144,0.026898667216300964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,2560,0.013812444276279874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,2560,0.012795555922720166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,2048,0.014584889014561972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,7168,0.02977777851952447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,3072,0.013742222554153867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,2048,0.01240711079703437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,1536,0.011807111402352652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,2048,0.014326221413082547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,3584,0.014728888869285583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,1536,0.01071999967098236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,1536,0.009696889254781935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,1536,0.012988444831636218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,2560,0.01568888955646091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,10240,0.05778577592637804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,768,0.008199999729792276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,1024,0.00832444429397583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,1024,0.011739555332395764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,512,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,768,0.007831999825106727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,2048,0.012143111063374413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,768,0.010724444356229572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,256,0.004571555389298333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,512,0.007381333245171442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,512,0.009032888544930352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,1024,0.009697777529557547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,128,0.0038728887836138406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,256,0.006646222124497096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,256,0.0069448889957533936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,256,0.008021333151393467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,64,0.003563555578390757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,128,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,128,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,32,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,128,0.008354666332403818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,768,0.009316444396972656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,64,0.006593777901596493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,32,0.008074666890833113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,512,0.007344889144102733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,65536,0.1748124493492974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,51200,0.21416000525156656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,65536,0.12838933202955458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,12288,0.04592355423503452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,1024,0.0107022225856781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,51200,0.10223022434446548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,16384,0.04809777935345968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,16384,0.0380311111609141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,12288,0.05526399943563673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,65536,0.27317688200208873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,65536,0.16336622503068712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,10240,0.04598133431540596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,51200,0.12835110558403864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,12288,0.03801511062516107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,16384,0.07111822234259711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,10240,0.03162755568822225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,8192,0.04016622238688999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,10240,0.033591111501057945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,8192,0.027268444498380024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,8192,0.02288355595535702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,7168,0.03401066528426276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,8192,0.030084444416893855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,16384,0.047817776600519814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,6144,0.029688000679016113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,12288,0.03643733263015747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,7168,0.024009777439965144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,7168,0.026659554905361597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,6144,0.02130399975511763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,10240,0.026035555534892615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,6144,0.02369600037733714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,5120,0.016532444291644625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,5120,0.018256000346607633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,4096,0.022619555393854778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,5120,0.020658666888872784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,12288,0.029688000679016113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,4096,0.01482311056719886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,3584,0.019194665882322524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,4096,0.018910222583346896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,6144,0.018240888913472492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,51200,0.13346044222513834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,3072,0.01700355609258016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,3584,0.017307556337780423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,3072,0.01315111087428199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,3072,0.016302223006884258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,4096,0.015524443652894763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,2560,0.011791999969217511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,2560,0.011740444435013665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,3584,0.014192000031471252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,2048,0.01237155579858356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,3072,0.012712889247470431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,2560,0.014260444376203748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,2048,0.014130666851997375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,1536,0.010558221903112201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,7168,0.020276443825827706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,1536,0.009456000394291347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,1536,0.008976000050703684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,1024,0.008504889077610439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,1536,0.012826667063766055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,5120,0.02568088968594869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,2560,0.014928888943460254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,1024,0.008061333662933773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,1024,0.008843555218643611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,1024,0.010995555255148145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,2048,0.010889777706729041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,512,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,768,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,768,0.009703111317422655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,512,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,256,0.0042364444169733255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,512,0.009187555147541894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,3584,0.01386755539311303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,768,0.0075297777851422625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,256,0.0069573331210348345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,128,0.0038702223036024305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,256,0.008223999705579545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,128,0.006657777975002925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,64,0.0035777779089079965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,128,0.00832799987660514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,32,0.003535110917356279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,512,0.0070186663005087115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,64,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,65536,0.20462577872806123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,32,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,256,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,65536,0.16548800468444824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,65536,0.1228684451844957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,2048,0.010024888647927178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,51200,0.1272302203708225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,51200,0.09823733568191528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,65536,0.14946488539377847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,16384,0.045479999648200146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,12288,0.04366755485534668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,16384,0.04650311006440056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,768,0.007810666329330868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,128,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,10240,0.03659910957018534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,12288,0.03461600012249417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,12288,0.03664088911480374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,16384,0.054696887731552124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,10240,0.030360000001059637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,8192,0.030751999881532457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,10240,0.0243493335114585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,10240,0.03215377860599094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,7168,0.026735110415352717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,8192,0.021456888980335657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,51200,0.11884800593058269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,16384,0.03593066665861342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,6144,0.02366222275627984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,7168,0.022659555077552795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,12288,0.029081778393851385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,6144,0.020072888996866014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,5120,0.020216888851589628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,8192,0.026178666286998328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,51200,0.15804978211720785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,8192,0.0277786652247111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,4096,0.017885333961910672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,5120,0.01590577761332194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,5120,0.02049777739577823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,4096,0.015049777097172208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,4096,0.014392889208263822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,6144,0.022614222433831956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,5120,0.017432888348897297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,3584,0.01719911065366533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,3072,0.013468444347381592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,3584,0.013213333156373767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,7168,0.02556444373395708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,3072,0.013040888640615674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,3584,0.01514400045077006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,2560,0.011745777395036487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,3072,0.01570577753914727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,4096,0.018058665924602084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,2048,0.010664000279373594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,2560,0.011296889020336999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,2560,0.011622221933470832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,2560,0.014667555689811707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,3584,0.014020444618331062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,1536,0.00868088917599784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,2048,0.009295110901196798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,2048,0.01370666672786077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,1024,0.007460444337791867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,1536,0.008952000074916417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,1536,0.012420444852775998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,6144,0.018349332941903006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,1024,0.008368888662921058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,1024,0.010370666782061258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,768,0.007334222396214803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,2048,0.010679999987284342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,768,0.009711999860074785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,768,0.007293333609898885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,7168,0.02015555567211575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,1536,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,256,0.0038844446341196695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,512,0.007068444457319047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,512,0.006329777754015393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,3072,0.012129777835475074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,768,0.005625777774386936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,256,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,256,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,128,0.00685777763525645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,128,0.008360000120268928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,64,0.0032195556494924757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,64,0.006264888991912206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,512,0.00868088917599784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,32,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,1024,0.00775911079512702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,65536,0.1783448855082194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,32,0.006635555376609166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,128,0.005969777703285217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,65536,0.12438489331139459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,512,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,65536,0.07212533553441365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,51200,0.0942302213774787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,16384,0.047008888589011304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,256,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,65536,0.13796977202097574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,51200,0.10957066880332099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,12288,0.03785066803296407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,16384,0.04257333278656006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,51200,0.1401146650314331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,12288,0.026877333720525105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,12288,0.01869155632125007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,10240,0.03125155634350247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,51200,0.0580728915002611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,128,0.003559111307064692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,10240,0.016674665941132438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,10240,0.0236124445994695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,10240,0.029947555727428857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,7168,0.023609777291615803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,8192,0.02612177696492937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,7168,0.017533333765135873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,12288,0.033891555335786604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,7168,0.013665777941544851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,6144,0.020581333173645865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,8192,0.027152000202072993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,6144,0.01276266657643848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,16384,0.03432710965474447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,5120,0.017631111873520743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,6144,0.021697777840826247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,5120,0.011639110743999481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,5120,0.014132445057233175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,5120,0.019328888919618394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,7168,0.023708444502618577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,4096,0.012520000338554382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,3584,0.013174222575293647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,4096,0.01738933391041226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,6144,0.016081778539551627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,8192,0.014862222803963555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,3584,0.011726222104496427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,3584,0.009725333087974125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,3584,0.016233777006467182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,4096,0.0148053334818946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,2560,0.010696888797812991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,3072,0.009098666409651438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,3072,0.010696888797812991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,3072,0.01533777846230401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,2048,0.008884444832801819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,2560,0.00832799987660514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,2560,0.00996888925631841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,4096,0.01075111081202825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,1536,0.008006222546100616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,2048,0.007815111014578078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,8192,0.020317332612143624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,2048,0.013373333546850415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,1024,0.005963555640644497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,1536,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,1536,0.011343999869293638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,1024,0.006618666566080517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,768,0.005240889059172736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,1024,0.009744000103738572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,2560,0.01402933398882548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,768,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,768,0.006579555571079254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,16384,0.023464888334274292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,3072,0.011775111158688864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,1536,0.007305777735180325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,256,0.003658666792843077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,512,0.006309333360857434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,512,0.008627555436558193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,1024,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,128,0.0033608888172441055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,256,0.006206222292449739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,256,0.008029333419269985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,2048,0.009229333036475712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,128,0.00600000015563435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,128,0.008023999631404877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,32,0.003212444484233856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,512,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,32,0.006174222048785951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,65536,0.1634275515874227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,256,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,65536,0.11945066187116836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,65536,0.06849066416422527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,512,0.004265777766704559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,51200,0.12780355082617864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,768,0.009418666362762451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,16384,0.04369066821204292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,51200,0.09295288721720378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,51200,0.056040002240075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,65536,0.13675110869937473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,16384,0.03394844465785556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,16384,0.022333333889643352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,51200,0.10937600003348456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,16384,0.04198844565285576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,128,0.005624889085690181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,12288,0.018087999688254464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,10240,0.028966221544477675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,12288,0.0336844457520379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,10240,0.01611111064751943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,64,0.005923555543025334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,8192,0.02509333358870612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,10240,0.029707554313871596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,8192,0.014222222897741528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,8192,0.020206222931543987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,7168,0.02162666618824005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,8192,0.02549244463443756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,12288,0.02589688864019182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,6144,0.018975999620225694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,7168,0.02345955537425147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,10240,0.022441777918073866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,6144,0.01575644479857551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,5120,0.016319110989570618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,6144,0.01236088905069563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,12288,0.035050666994518705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,5120,0.011196444431940714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,4096,0.014208000567224292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,7168,0.017292444904645283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,4096,0.010063110954231685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,4096,0.01257333325015174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,3584,0.012528888881206512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,6144,0.020928000410397846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,3584,0.011510222322411008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,3584,0.009184888667530483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,3584,0.01605866683853997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,5120,0.013929777675204806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,64,0.0029253332565228143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,2560,0.010031999813185798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,3072,0.010767111347781287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,3072,0.008427555362383524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,3072,0.01532977819442749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,2560,0.007911110917727152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,2560,0.01387555566098955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,3072,0.011342222491900126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,2048,0.008366222182909647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,1536,0.007688889073001013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,7168,0.013447110851605734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,1536,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,1536,0.011054221954610614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,1024,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,1024,0.007295110987292395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,2048,0.013012444807423485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,2048,0.007485333416197035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,768,0.004934222333961063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,1024,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,1024,0.009363555245929295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,4096,0.017059556312031217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,768,0.006933333145247565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,2048,0.008619555168681675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,768,0.008994667066468133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,512,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,256,0.003547555456558863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,512,0.008607111043400234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,5120,0.01902577777703603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,256,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,256,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,128,0.0032026666320032547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,256,0.008015111088752747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,2560,0.009733333355850643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,1536,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,512,0.003962666624122196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,128,0.006042666733264923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,32,0.0031893334041039148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,128,0.008024888734022776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,65536,0.13248088624742296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,512,0.006015111174848344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,64,0.005776888794369168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,65536,0.11373689439561631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,65536,0.06621778011322021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,51200,0.10530222124523586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,768,0.006614222294754452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,51200,0.05404444535573324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,128,0.005487111293607288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,51200,0.10949599742889404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,16384,0.03240799903869629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,32,0.0063688887490166565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,12288,0.034325334760877825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,16384,0.04161599940723843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,12288,0.01756355497572157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,64,0.0029226665695508323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,65536,0.1362480057610406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,12288,0.02511022157139248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,12288,0.03343822227583991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,10240,0.02163200080394745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,16384,0.021015110943052504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,8192,0.01919644408755832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,51200,0.08767377667956883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,8192,0.013911111487282647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,7168,0.01999022232161628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,10240,0.027443556321991816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,8192,0.02555199960867564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,7168,0.01291822228166792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,7168,0.016875555117925007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,6144,0.01723911033736335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,10240,0.02915289004643758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,6144,0.011923555698659686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,6144,0.015211555692884656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,5120,0.015233778291278414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,6144,0.020986666282018025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,5120,0.010701333483060202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,4096,0.013108444710572561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,5120,0.018696000178654987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,10240,0.01585422290696038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,4096,0.016843555702103507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,4096,0.009386666946940953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,3584,0.011312888728247749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,16384,0.04153155618243747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,3584,0.008664888640244802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,3072,0.010780444575680627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,3584,0.015848888291252982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,3584,0.0119982221060329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,3072,0.010434666441546546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,2560,0.009729777773221334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,3072,0.014870221416155497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,2560,0.008996444443861643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,2560,0.008052444292439355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,2048,0.008456889126035903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,2560,0.013456000222100152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,2048,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,8192,0.022031999296612207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,2048,0.01236177815331353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,3072,0.008365333080291748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,1536,0.011058666639857821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,1024,0.005478222337034013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,5120,0.013728888498412238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,1024,0.006979555719428592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,1024,0.009641778137948778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,1024,0.006643555644485686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,1536,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,768,0.006614222294754452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,1536,0.006980444822046492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,768,0.006455111006895701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,768,0.009015999734401703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,512,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,7168,0.023100444012218054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,768,0.004949333353175057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,512,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,256,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,256,0.007967111137178209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,512,0.0042204442951414315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,256,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,128,0.005485333502292633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,1536,0.0075502221783002215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,128,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,512,0.00868977771864997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,32,0.0032115555885765287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,64,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,32,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,128,0.0031795555518733132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,65536,0.1078888906372918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,256,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,128,0.006008000009589725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,51200,0.09285422166188557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,65536,0.13674044609069824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,51200,0.08486488792631362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,65536,0.11518222755855984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,16384,0.03581422236230638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,51200,0.1088800032933553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,16384,0.0310479998588562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,4096,0.012132444315486483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,12288,0.03012977706061469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,64,0.0031964443624019623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,16384,0.04148088892300924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,10240,0.027099554737408955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,12288,0.016872000363137987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,12288,0.023797333240509033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,12288,0.03261333372857835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,8192,0.018960000740157235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,10240,0.020848888489935134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,10240,0.02864711152182685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,8192,0.013453333742088742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,7168,0.016600888636377122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,16384,0.020618667205174763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,65536,0.06395555867089166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,7168,0.022775999373859827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,10240,0.015218666858143277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,6144,0.014741332994567024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,6144,0.014952000644471912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,6144,0.020615999897321064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,8192,0.025241777300834656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,7168,0.016287111573749118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,5120,0.01351999988158544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,7168,0.012482666307025485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,51200,0.051876445611317955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,4096,0.011761777930789523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,5120,0.01847466660870446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,4096,0.011731555892361535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,3584,0.011385777758227454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,5120,0.012753778033786349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,3584,0.010977778169843884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,3584,0.015437333120240105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,3584,0.008664888640244802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,3072,0.010055999788973067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,8192,0.01811288959450192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,3072,0.009810667071077559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,2048,0.008605333666006723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,6144,0.011501332951916588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,3072,0.00814488861295912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,3072,0.014803555276658801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,2560,0.008663111262851292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,2560,0.007640889121426477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,2560,0.013454222016864352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,5120,0.010094222095277574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,2048,0.007349333001507654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,1536,0.0069919998447100324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,2048,0.011748444702890186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,4096,0.016720000240537856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,2560,0.00927466650803884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,1024,0.005375111268626318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,1536,0.010447111394670276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,2048,0.008000888758235509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,1024,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,768,0.004937777916590373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,1024,0.009391110804345872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,2048,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,4096,0.008777778181764815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,512,0.004238222208287981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,768,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,768,0.009011555876996782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,256,0.003540444291300244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,512,0.006162666612201267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,512,0.005902222047249476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,512,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,128,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,256,0.005964444329341252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,256,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,1024,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,64,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,128,0.007715555528799693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,768,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,64,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,1536,0.0070088886552386815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,65536,0.09204266468683879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,256,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,65536,0.06861066818237305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,51200,0.0752684473991394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,65536,0.06155021985371908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,65536,0.13600977261861166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,32,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,51200,0.052758223480648465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,1536,0.0069955554273393415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,51200,0.10848000314500596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,32,0.005848000033034219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,12288,0.0258906665775511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,16384,0.019938665959570143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,16384,0.040940443674723305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,12288,0.016190222567982145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,12288,0.01628444426589542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,16384,0.030228445927302044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,51200,0.049560887946022876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,10240,0.014644443988800049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,16384,0.020563556088341605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,10240,0.028200000524520874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,128,0.005995555470387141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,10240,0.02259911100069682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,7168,0.014521777629852295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,8192,0.013017777767446307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,10240,0.01460088955031501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,7168,0.011787555283970304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,6144,0.012793777717484368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,8192,0.01626755628320906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,7168,0.02275111112329695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,128,0.005241777747869492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,8192,0.013444444371594323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,8192,0.024785778588718835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,5120,0.010097777677906884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,5120,0.018343110879262287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,4096,0.010085333552625444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,7168,0.011894221935007306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,4096,0.009191999832789103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,12288,0.03228622343805101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,3584,0.008999110923873054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,3584,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,5120,0.009504000345865885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,3584,0.008218666745556725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,3584,0.01532977819442749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,6144,0.010900444454616971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,4096,0.009032888544930352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,3072,0.007627555893527136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,2560,0.007385777930418651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,3072,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,3072,0.014540443817774454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,2560,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,2560,0.0075146663520071245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,2048,0.006608888920810487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,3072,0.008628444539176093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,2048,0.006675555474228329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,6144,0.011050666371981302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,1536,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,1536,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,1536,0.010402667025725046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,1536,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,1024,0.004574222283230888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,2560,0.0127688886390792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,1024,0.005979555762476391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,2048,0.011742221812407175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,768,0.003906666818592283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,1024,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,768,0.0058808889653947614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,768,0.00903733323017756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,512,0.003591999825504091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,5120,0.011720000041855706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,1024,0.009372444616423713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,512,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,256,0.0035662220584021676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,512,0.008348444269763099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,768,0.006267555471923616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,256,0.007877333296669854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,256,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,512,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,128,0.005236444373925527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,64,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,128,0.005990222096443176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,32,0.0028880000528362063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,64,0.005274666680230035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,256,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,65536,0.08267644378874037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,128,0.0031884445084465873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,6144,0.02031822171476152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,65536,0.06437333424886067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,65536,0.060580445660485156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,65536,0.13667910628848606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,128,0.007689778175618913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,51200,0.05081955591837565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,16384,0.028223110569847956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,51200,0.048624889718161694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,51200,0.10825422075059678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,12288,0.023362666368484497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,16384,0.019644444187482197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,16384,0.04064088728692796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,4096,0.016566221912701923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,12288,0.015827554795477126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,51200,0.06712977753745185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,12288,0.03233333428700765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,10240,0.014504889647165934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,8192,0.018903111418088276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,10240,0.02811466654141744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,16384,0.019952888290087383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,8192,0.012757333616415659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,7168,0.014004444082578024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,8192,0.02455377744303809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,8192,0.012444444828563266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,10240,0.021360889077186584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,6144,0.012429333395428128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,7168,0.011510222322411008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,7168,0.022413333257039387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,10240,0.014113777213626437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,5120,0.011046222514576383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,6144,0.010391111175219217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,6144,0.020432889461517334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,32,0.005419555637571547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,5120,0.009687999884287516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,4096,0.010056888891590966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,5120,0.018223111828168232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,2048,0.007080000307824876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,3584,0.009043555292818282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,4096,0.00868088917599784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,4096,0.016559110747443307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,3584,0.007705777883529663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,3584,0.008311111066076491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,3072,0.008371555142932469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,5120,0.009826666778988307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,3072,0.013807999591032663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,2560,0.007353777686754863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,12288,0.016184000505341422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,4096,0.008391111261314815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,2560,0.006851555572615729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,2048,0.006319111006127463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,2560,0.012429333395428128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,3584,0.01495733360449473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,2048,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,1536,0.00517155561182234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,2048,0.011412444214026133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,3072,0.008006222546100616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,7168,0.011420444481902652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,1536,0.006865777903132969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,1536,0.010384889112578498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,2560,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,6144,0.010808000134097205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,1024,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,768,0.0038853333228164246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,1024,0.009364444348547194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,2048,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,3072,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,768,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,768,0.009119999905427298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,1536,0.005992888990375731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,512,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,256,0.0032595555401510666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,256,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,256,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,256,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,128,0.0028880000528362063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,768,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,128,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,512,0.008360889222886827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,64,0.00318755561278926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,128,0.005968889014588461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,32,0.002879111096262932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,1024,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,65536,0.07229777839448717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,32,0.005551110953092575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,65536,0.0627760026190016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,512,0.005631111148330901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,128,0.007684444387753804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,51200,0.057164443863762744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,51200,0.04911555515395271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,512,0.003587555554178026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,51200,0.04752622379197014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,16384,0.02251022226280636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,1024,0.004223111189073986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,16384,0.01923022170861562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,65536,0.05952533086140951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,12288,0.019169777631759644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,12288,0.015271999769740634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,64,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,12288,0.03181333343187968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,51200,0.10849066575368245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,10240,0.013904000322024027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,8192,0.01499466598033905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,10240,0.02819377846188015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,65536,0.13606933752695718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,16384,0.04061244593726264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,8192,0.012528888881206512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,8192,0.011767999993430244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,12288,0.01586577792962392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,10240,0.01663644446267022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,6144,0.011723555624485016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,7168,0.02239111065864563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,10240,0.014289778139856128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,6144,0.010363555616802638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,5120,0.010475555227862464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,6144,0.010085333552625444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,6144,0.020340444313155282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,16384,0.019241778386963736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,7168,0.013125333521101209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,5120,0.009356444080670675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,5120,0.009413333402739631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,5120,0.018265777164035372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,3584,0.008645333349704742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,7168,0.011128000087208219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,4096,0.015784000356992085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,3584,0.007676444119877285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,3584,0.014488000008794995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,4096,0.009684444301658207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,8192,0.024383111132515803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,3072,0.008008889026112026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,4096,0.00832799987660514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,3072,0.013457777599493662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,3584,0.00816000004609426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,3072,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,2560,0.012081777883900536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,2048,0.0064071110553211635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,2560,0.0074177781740824384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,3072,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,2048,0.00646577775478363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,4096,0.008616000413894653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,1536,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,1536,0.005961777849329843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,1536,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,1536,0.010313777459992303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,1024,0.00583022211988767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,1024,0.009364444348547194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,1024,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,7168,0.010769777827792697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,768,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,1024,0.0042275554604000514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,768,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,2560,0.006959111326270633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,768,0.008986666798591614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,512,0.005991111199061076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,256,0.0033324443631701996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,2048,0.0070168889231152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,768,0.0038986665507157645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,256,0.005917333480384614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,256,0.004951999833186467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,256,0.007728888756699032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,512,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,128,0.004931555440028508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,64,0.0029377777957253983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,128,0.005919111271699269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,128,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,32,0.002874666617976295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,512,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,65536,0.062457780043284096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,512,0.008379555410808986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,64,0.005223999834722943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,32,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,2048,0.011115555961926779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,128,0.0028702221396896574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,65536,0.13581777943505183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,51200,0.04650400082270304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,16384,0.021962665849261816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,51200,0.04723199870851305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,51200,0.10858933130900066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,51200,0.04896177848180135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,16384,0.019256000717480976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,16384,0.04026133484310574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,12288,0.015114666687117683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,16384,0.018562666243977018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,12288,0.015588444140222339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,12288,0.01919644408755832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,10240,0.013568888935777875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,10240,0.013491555220550962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,10240,0.028011557128694322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,8192,0.014512888259357877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,12288,0.03214577833811442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,8192,0.011398221883508893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,7168,0.012888888518015543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,8192,0.024480889240900677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,2560,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,6144,0.011520889070298938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,7168,0.011130666567219628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,7168,0.02215822206603156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,6144,0.010076444182131026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,6144,0.010391111175219217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,6144,0.020260444945759244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,10240,0.01629244453377194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,5120,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,4096,0.008948444492287105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,5120,0.009036444127559662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,65536,0.0583404435051812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,7168,0.011076444553004371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,3584,0.008292444050312042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,4096,0.008394666843944127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,4096,0.00795288880666097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,4096,0.015784889459609985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,3072,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,3584,0.007424888511498769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,3584,0.014502222339312235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,8192,0.011771555576059552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,3072,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,2560,0.007086222370465596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,3072,0.013122667041089801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,65536,0.060432891050974526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,2048,0.00600622221827507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,2560,0.006852444261312485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,2560,0.012111999922328524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,2048,0.006412444429265128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,5120,0.010029333333174387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,2048,0.01147377739350001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,3072,0.007167999943097432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,1536,0.006005333529578314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,1024,0.003908444609906939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,1536,0.010724444356229572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,5120,0.017975999249352347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,2560,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,768,0.00388711111413108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,1024,0.006528000036875407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,1024,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,1024,0.009407111340098912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,512,0.003567111161020067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,768,0.005640888793600931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,768,0.008904000123341879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,512,0.0053004444473319584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,512,0.005975111077229182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,512,0.008346666892369589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,1536,0.0069511110583941145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,256,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,256,0.0052062223354975385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,256,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,3584,0.008052444292439355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,1536,0.0048942222363419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,64,0.0028764444092909494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,128,0.007731555236710443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,32,0.0028942223224374983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,256,0.0032195556494924757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,64,0.004973333328962326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,32,0.005305777821275923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,128,0.003212444484233856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,65536,0.05783466498057047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,51200,0.04259555538495382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,128,0.005034666508436203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,128,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,51200,0.045296000109778516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,16384,0.01847111185391744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,65536,0.05224622289339701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,51200,0.10825866460800171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,16384,0.01811377704143524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,12288,0.015227556228637695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,16384,0.04026222229003906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,2048,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,65536,0.1357146633995904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,12288,0.01475911173555586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,10240,0.01458755632241567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,12288,0.03182755576239692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,51200,0.04532000091340807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,8192,0.012817777693271637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,10240,0.012807110945383707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,10240,0.027793778313530818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,16384,0.018783110711309645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,8192,0.011629333098729452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,7168,0.012084444363911947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,768,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,12288,0.014101333088344999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,7168,0.010727110836240979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,7168,0.01039644413524204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,7168,0.022081777453422546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,10240,0.013159111142158508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,6144,0.009728000395827824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,5120,0.012763555679056378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,8192,0.02440622283352746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,65536,0.056629333231184215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,5120,0.008667555948098501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,4096,0.010879110958841113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,5120,0.01723911033736335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,6144,0.0111191107167138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,4096,0.008341333104504479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,3584,0.009296000003814697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,6144,0.009356444080670675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,6144,0.019813333948453266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,3072,0.008369777765538957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,3584,0.008079110748238033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,3584,0.014152000347773233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,5120,0.009120889008045197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,3072,0.007672000262472365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,3072,0.01313777764638265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,4096,0.007650666766696506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,2560,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,4096,0.015495111544926962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,2560,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,3584,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,2048,0.0063999998900625445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,1536,0.004888888862397936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,2048,0.011120888921949597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,3072,0.00700711127784517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,1536,0.005997333261701796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,1536,0.0069448889957533936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,1024,0.0042151109211974675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,2048,0.00674311113026407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,2560,0.012186666329701742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,768,0.003863999827040566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,1024,0.006312888943486744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,2048,0.0069982219073507525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,2560,0.0073315559162033936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,768,0.00870666652917862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,512,0.0035413333939181436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,1536,0.010115555591053432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,512,0.005299555758635203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,512,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,8192,0.011337777806652917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,1024,0.005771555420425203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,1024,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,128,0.0031893334041039148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,256,0.005807111246718301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,256,0.00775999989774492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,128,0.004976888911591636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,64,0.0028888889484935333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,128,0.007903999752468532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,32,0.0028862222615215513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,256,0.0032151111712058387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,65536,0.03959555427233378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,32,0.005053333524200651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,256,0.004943111290534337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,768,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,65536,0.0553342236412896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,51200,0.03168355425198873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,128,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,51200,0.04389866524272495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,16384,0.01774044500456916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,51200,0.10769955317179362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,512,0.008307555483447181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,16384,0.0181057784292433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,16384,0.01755288905567593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,16384,0.04011911153793335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,65536,0.055571556091308594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,768,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,10240,0.013408000270525614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,12288,0.031593776411480375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,51200,0.04478755593299866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,64,0.0052675555149714155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,10240,0.01278222186697854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,10240,0.027790221903059218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,12288,0.01463822192615933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,8192,0.01107377807299296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,7168,0.01110666659143236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,8192,0.0236799998415841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,12288,0.013471110827393003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,7168,0.010058666268984476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,6144,0.011056889262464313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,7168,0.010383110907342698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,7168,0.021299555897712708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,8192,0.012474666866991254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,5120,0.010000000397364298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,6144,0.009706666900051964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,6144,0.01891644464598762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,8192,0.01105955574247572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,5120,0.009332444104883406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,65536,0.13583377997080484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,4096,0.007648000286685095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,4096,0.015153777268197803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,4096,0.008325333396593729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,3584,0.007305777735180325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,6144,0.009313777916961247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,5120,0.00869511150651508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,3584,0.0074720001882976955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,3072,0.006665777828958299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,3584,0.008018666671382057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,3072,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,10240,0.0124177775449223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,3072,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,2560,0.006706666615274217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,4096,0.009326222042242685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,3584,0.014382221632533602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,2048,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,2560,0.012098666694429187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,2048,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,1536,0.004592000196377437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,2048,0.007002666592597961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,2048,0.011063111325105032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,1536,0.006743999818960826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,1024,0.0038764443662431505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,5120,0.017198221551047433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,2560,0.007419555551475949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,1024,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,768,0.0039022221333450745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,3072,0.013439111411571503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,2560,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,768,0.005567111074924469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,512,0.003256888853179084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,768,0.009043555292818282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,1536,0.006127111199829314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,1536,0.01037333326207267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,12288,0.014187556174066333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,256,0.0031831111345026228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,512,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,512,0.008384000096056197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,1024,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,256,0.005160888863934411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,256,0.008035555481910706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,128,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,128,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,64,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,128,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,32,0.002528888897763358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,65536,0.02772444486618042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,64,0.00498488876554701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,512,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,65536,0.05270933442645603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,65536,0.05482666691144308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,51200,0.02437066700723436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,65536,0.13577422830793592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,51200,0.04213688770929972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,51200,0.04455288913514879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,51200,0.10745866431130303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,32,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,16384,0.03985155622164408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,12288,0.012616888516479068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,1024,0.009082666701740688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,12288,0.013800889253616333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,768,0.006275555739800136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,16384,0.013784888717863293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,12288,0.013526221944226159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,16384,0.017238222890430026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,10240,0.011911110745535957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,8192,0.01311822235584259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,10240,0.02712533374627431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,16384,0.016548444827397663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,8192,0.011066666907734342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,8192,0.010424888796276517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,8192,0.023062222533755835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,12288,0.030919111437267725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,7168,0.010038221875826517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,7168,0.010374222364690568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,7168,0.021311110920376245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,256,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,6144,0.018575110369258456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,5120,0.010022222167915767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,10240,0.011216888825098673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,5120,0.008319999608728621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,6144,0.010388444695207808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,4096,0.008802666432327693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,5120,0.016901334126790363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,128,0.0029164445069101122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,4096,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,3584,0.007115555306275685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,4096,0.008384000096056197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,4096,0.014925332532988654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,3584,0.008025777836640676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,3584,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,3584,0.014180445008807711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,5120,0.009018667042255402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,3072,0.006980444822046492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,2560,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,10240,0.012415111064910889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,6144,0.009689778089523315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,7168,0.012274666792816587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,2048,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,2560,0.007476444045702617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,2560,0.01181422256761127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,3072,0.006599999964237213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,2048,0.007112888826264276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,2048,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,1536,0.005579555614127054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,2048,0.01108000013563368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,3072,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,3072,0.013481777575280933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,1536,0.010399999717871347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,2560,0.006659555352396435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,1024,0.005649777750174205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,768,0.004253333227501975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,1024,0.009408888717492422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,768,0.005321777943107817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,1536,0.006674666785531574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,768,0.00906044410334693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,512,0.0035422220826148987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,1024,0.0062977779242727495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,512,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,256,0.004943999979231092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,6144,0.009020444419648912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,256,0.0057813334796163775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,256,0.007707555260923173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,512,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,512,0.008291555775536431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,128,0.004962666581074397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,128,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,128,0.008031999899281396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,32,0.0029004443850782183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,64,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,768,65536,0.029845333761639063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,128,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,65536,0.05168177684148153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,64,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,65536,0.13542400466071233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,1536,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,51200,0.04172888729307386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,32,0.004965333475006951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,768,16384,0.011951111257076263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,256,0.0032542221662071017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,65536,0.05458133419354757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,16384,0.016399999459584553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,768,12288,0.010563555690977307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,16384,0.03947110970815023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,1024,0.004588444613748127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,12288,0.013223110801643796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,768,10240,0.009742221898502773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,12288,0.030886222918828327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,51200,0.10733422305848862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,10240,0.011792889071835412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,10240,0.01221422188811832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,10240,0.026761778526835974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,16384,0.016188444362746347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,768,0.006341333190600078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,8192,0.01106844428512785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,8192,0.02285066743691762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,12288,0.013445333474212222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,51200,0.043980442815356784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,7168,0.010384000009960597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,7168,0.009716444545321995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,7168,0.021086222595638696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,768,8192,0.010127999716334874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,6144,0.009728888670603434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,6144,0.018626666731304593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,8192,0.010401777923107147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,768,7168,0.00870488915178511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,768,4096,0.007691555553012424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,5120,0.016869333055284288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,768,6144,0.009370666411187913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,4096,0.007691555553012424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,4096,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,768,3584,0.007666666474607255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,768,51200,0.02420088814364539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,3584,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,768,3072,0.006437333507670297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,5120,0.008448888858159384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,5120,0.009036444127559662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,3072,0.007179555793603261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,3072,0.007666666474607255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,3072,0.013414222333166333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,4096,0.015037332971890768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,2560,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,2560,0.007415999968846639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,2560,0.012226666841242047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,3584,0.007984889050324758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,3584,0.014198222094111972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,768,1536,0.004611555486917496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,2048,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,2048,0.01144266708029641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,768,2560,0.005920889063013925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,768,1024,0.0038817777401871155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,1536,0.010214221974213919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,6144,0.009009777671760982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,768,2048,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,768,768,0.003908444609906939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,1024,0.0063271112740039825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,2048,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,768,0.0053102220926019884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,768,512,0.0032177778581778207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,768,0.008681777450773451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,1536,0.005988444305128521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,1536,0.006660444455014334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,512,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,512,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,768,256,0.0032551110618644287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,512,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,1024,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,768,128,0.0028400001012616684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,256,0.005655111124118169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,256,0.004973333328962326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,256,0.007686222592989604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,768,64,0.0028435554769304064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,128,0.005640888793600931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,128,0.005085333353943295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,768,32,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,128,0.007635555333561367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,768,0.006275555739800136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,64,0.005008888741334279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,32,0.004899555610285865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,512,51200,0.018238221605618794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,65536,0.05065777897834778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,65536,0.054173333777321704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,65536,0.13545244269900852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,512,16384,0.012120000190205045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,1024,0.009350222018029954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,51200,0.10708089007271661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,16384,0.016116445263226826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,16384,0.016186666157510545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,12288,0.012871111432711283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,768,5120,0.007696888513035244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,12288,0.01349955548842748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,512,10240,0.011063111325105032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,12288,0.03019022279315525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,10240,0.012101333174440594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,512,12288,0.01070844464831882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,10240,0.011744000017642975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,10240,0.02679377794265747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,8192,0.01107911103301578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,512,7168,0.008797333472304875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,8192,0.023430221610599097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,51200,0.04176622298028734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,7168,0.009715555442704095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,512,6144,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,512,65536,0.023357333408461675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,512,8192,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,6144,0.009114666945404476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,512,5120,0.008004444340864817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,6144,0.018382221460342407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,51200,0.040977779361936785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,5120,0.008448888858159384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,512,4096,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,5120,0.00904977818330129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,7168,0.010384889112578498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,16384,0.03841244512134128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,4096,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,4096,0.01515111161602868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,6144,0.009704888694816166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,8192,0.01038933296998342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,512,3072,0.006331555545330048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,3584,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,3584,0.013908444179428948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,3072,0.006950221955776215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,3072,0.007636444436179266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,3072,0.01276622215906779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,7168,0.021021333005693223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,4096,0.008366222182909647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,2560,0.006602666858169768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,512,2048,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,3584,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,2048,0.01144177797767851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,512,1536,0.004549333204825719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,512,2560,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,1536,0.005982222242487802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,1536,0.010043555663691627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,2560,0.012105777859687805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,512,1024,0.003986666599909465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,1536,0.006793777975771163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,2048,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,1024,0.0063493334584765965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,512,768,0.003588444242874781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,2048,0.00703022215101454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,5120,0.016728889611032274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,768,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,512,512,0.003370666669474708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,768,0.009033777647548253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,512,3584,0.0069511110583941145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,512,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,512,0.005290666802061929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,512,0.008013333711359236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,256,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,512,128,0.0028728888266616394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,256,0.007697777615653143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,1024,0.009332444104883406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,128,0.005970666805903117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,128,0.004907555464241239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,512,64,0.0028657778683635923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,128,0.008016889293988546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,2560,0.007344889144102733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,512,256,0.0031840000301599503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,256,65536,0.016436444388495553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,32,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,256,0.005087999833954705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,65536,0.050082667006386646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,256,51200,0.014527110589875115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,65536,0.13475467099083796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,768,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,51200,0.04016088777118259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,256,16384,0.009696889254781935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,51200,0.10648622115453084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,64,0.004945777770545748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,16384,0.016203555795881484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,256,12288,0.008325333396593729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,65536,0.04709955718782213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,16384,0.03879822293917338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,12288,0.01348622226052814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,12288,0.012808000048001608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,12288,0.030547555949952867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,512,32,0.002576888849337896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,51200,0.0380968881977929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,256,8192,0.008018666671382057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,10240,0.011533333195580376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,10240,0.02676888969209459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,16384,0.015622221761279635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,256,7168,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,1024,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,8192,0.02269511090384589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,7168,0.010218666659461128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,7168,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,256,6144,0.007663110891977946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,7168,0.020623111062579684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,256,5120,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,6144,0.009712888962692684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,10240,0.012355555262830524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,6144,0.018417777286635507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,256,4096,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,5120,0.009012444151772393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,5120,0.008356444537639618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,8192,0.01070577816830741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,5120,0.016546666622161865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,256,3584,0.0070426662762959795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,4096,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,4096,0.007442666424645319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,256,3072,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,4096,0.01516177753607432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,256,10240,0.00903200027015474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,3584,0.007256888680987888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,256,2560,0.006783111227883234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,3072,0.007681777907742395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,3072,0.0069751110341813825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,6144,0.009026666482289633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,256,2048,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,2560,0.011786667009194693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,8192,0.011028444601429833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,2048,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,2048,0.007017778025733099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,256,1536,0.004548444516128964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,2048,0.011445333560307821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,3584,0.014167999227841696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,1536,0.006648889018429651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,1536,0.010094222095277574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,2560,0.007272889216740926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,1024,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,256,768,0.003568888952334722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,1024,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,3584,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,768,0.005436444448100195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,256,512,0.0032257777121331957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,256,1024,0.003882666842805015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,3072,0.013331555657916598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,512,0.005304000029961268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,2560,0.006618666566080517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,256,256,0.003117333269781537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,512,0.008667555948098501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,256,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,256,128,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,256,0.007680888805124495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,768,0.00903555585278405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,1536,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,256,64,0.002571555475393931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,128,0.008012444608741337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,256,32,0.002546666603949335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,512,0.00592533333433999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,128,65536,0.015535111228624979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,64,0.004993777722120285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,32,0.005013333426581489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,256,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,128,51200,0.013153778182135688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,65536,0.04798488815625509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,65536,0.13231999344295925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,128,0.004857777721352047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,51200,0.038196444511413574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,51200,0.03770844472779168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,51200,0.10524444447623359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,16384,0.016176889340082806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,16384,0.01579466627703773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,16384,0.03849600089920892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,65536,0.04662755462858412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,768,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,12288,0.030177778667873804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,128,10240,0.007672888537247975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,128,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,10240,0.012100444071822695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,10240,0.02660177813635932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,1024,0.00630044440428416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,128,8192,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,8192,0.010286221901575724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,8192,0.023009777069091797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,12288,0.012981333666377597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,128,7168,0.006950221955776215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,128,16384,0.009402666654851701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,7168,0.010285333626800114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,10240,0.011732444167137146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,128,6144,0.006656888872385025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,128,12288,0.008347555167145198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,6144,0.009702222214804756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,128,5120,0.006320000108745363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,6144,0.018568888306617737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,12288,0.01350400017367469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,5120,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,128,4096,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,5120,0.0090506664580769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,7168,0.009683555199040307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,7168,0.020968889196713764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,128,3584,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,4096,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,4096,0.015084443820847405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,6144,0.009019555317031013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,128,3072,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,3584,0.008060444560315874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,3584,0.014135110709402295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,8192,0.01072266697883606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,3072,0.006979555719428592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,5120,0.016710221767425537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,2560,0.006659555352396435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,128,2048,0.005976000179847081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,4096,0.008325333396593729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,2560,0.012113778127564324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,2048,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,3584,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,2048,0.011471110913488599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,128,2560,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,1536,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,128,1024,0.0038986665507157645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,2560,0.007299555672539606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,1536,0.010398222340477837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,1024,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,128,768,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,1024,0.009359111388524374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,2048,0.006989333364698622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,128,1536,0.004279110994603899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,128,512,0.0032186667538351486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,768,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,3072,0.007697777615653143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,768,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,3072,0.013103110922707452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,128,256,0.0029173334025674393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,512,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,512,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,256,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,256,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,1024,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,128,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,128,64,0.002554666664865282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,128,0.007648888561460707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,128,32,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,1536,0.006682666639486949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,512,0.005985777825117111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,64,65536,0.014106666048367819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,32,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,64,51200,0.012110222544935016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,64,16384,0.007984889050324758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,256,0.007663110891977946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,65536,0.047304888566335045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,64,12288,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,64,10240,0.007296889192528195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,16384,0.015676443775494892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,64,8192,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,768,0.005552000055710475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,64,7168,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,10240,0.011415999796655444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,64,6144,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,8192,0.01037599974208408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,64,5120,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,7168,0.009706666900051964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,128,128,0.0028515555378463534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,6144,0.009009777671760982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,51200,0.03810844487614102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,64,3584,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,128,0.0047493332790003884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,12288,0.012824888858530255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,64,2560,0.006302222195598815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,64,0.005249777601824866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,64,2048,0.0061048890153567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,64,4096,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,64,1536,0.004573333180612988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,2560,0.006894222150246303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,5120,0.008303110798199972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,2048,0.0063484443558586975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,64,768,0.0035377778112888336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,3584,0.007293333609898885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,64,512,0.003306666596068276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,768,0.005493333356248007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,512,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,64,256,0.0028560000161329904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,64,1024,0.003951999876234266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,4096,0.0074720001882976955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,256,0.005009777843952179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,64,64,0.0028488888508743713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,1024,0.005625777774386936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,64,32,0.002528888897763358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,64,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,32,65536,0.012430222498046027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,32,0.00490488898422983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,32,51200,0.011061333119869232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,65536,0.04723199870851305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,32,16384,0.007072000039948358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,1536,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,51200,0.038013332419925265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,32,12288,0.007661333514584436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,32,10240,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,3072,0.006704888823959563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,32,8192,0.00675644435816341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,12288,0.012774222426944308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,32,7168,0.0069475554757648045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,10240,0.011408000356621213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,32,6144,0.006607111129495833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,8192,0.010046222143703038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,32,5120,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,7168,0.009701333112186855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,32,4096,0.00591822216908137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,64,3072,0.006129777679840724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,6144,0.009012444151772393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,32,3584,0.006274666637182236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,32,3072,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,4096,0.007623111208279927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,128,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,3584,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,3072,0.006792888873153263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,32,1536,0.004294222427739037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,2560,0.006688000013430913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,32,1024,0.003882666842805015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,1536,0.005976888868543837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,32,768,0.0033724444607893624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,1024,0.005663110978073544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,5120,0.00812355594502555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,32,512,0.003245333416594399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,32,256,0.0028880000528362063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,16384,0.015832000308566622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,32,2048,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,512,0.005085333353943295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,32,64,0.002520888836847411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,256,0.005053333524200651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,32,32,0.002577777744995223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,2048,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,64,0.004823110997676849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,64,128,0.002620444529586368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,32,0.004955555415815777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,768,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,32,2560,0.006372444331645966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,96,32,128,0.0028391112056043413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,16384,0.37111377716064453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,51200,0.6052311261494955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,12288,0.27855287657843697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,128,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,16384,0.13295643859439427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,12288,0.14841066466437444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,10240,0.20743821726904976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,12288,0.10031021965874566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,10240,0.08454844686720107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,51200,0.40880799293518066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,7168,0.15076266394721136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,8192,0.07711466815736559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,6144,0.12883822123209634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,51200,1.1376667022705078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,7168,0.06150933106740316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,7168,0.0942115585009257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,5120,0.10945600271224976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,6144,0.05448177787992689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,6144,0.07854755719502766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,10240,0.1287626690334744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,5120,0.06601688596937391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,3584,0.07923466629452176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,4096,0.03940444522433811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,4096,0.09193866782718235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,3072,0.06864800055821736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,3584,0.04099911120202806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,3584,0.04832355512513054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,3072,0.036052445570627846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,8192,0.17269510693020293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,5120,0.04618044363127815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,2048,0.04716622167163425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,2560,0.036673777633243136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,2560,0.027357333236270483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,8192,0.10632177856233384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,1536,0.03732622332043118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,2048,0.026495110657480028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,2048,0.03187555405828688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,1024,0.02823822365866767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,1536,0.018611555298169453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,768,0.020975111259354487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,1024,0.016878222425778706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,1024,0.02944444285498725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,3072,0.04222311245070564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,512,0.014671110444598727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,2560,0.05796000030305651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,768,0.022036444809701707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,256,0.009336888790130615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,512,0.012784889174832238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,4096,0.054208887947930225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,128,0.00833422193924586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,256,0.011076444553004371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,256,0.012479999827014076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,16384,0.20134133762783477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,64,0.0059262220230367445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,128,0.011319110790888468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,32,0.006335999816656113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,128,0.011378666592968835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,64,0.011405333048767514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,768,0.013602666556835175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,512,0.0165928883685006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,65536,0.6230897903442383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,65536,1.0938355127970378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,1536,0.04108355442682902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,51200,0.8589795430501302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,16384,0.2798355685340034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,32,0.015192889504962497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,65536,0.416885322994656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,16384,0.15858222378624812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,16384,0.10694044166141087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,12288,0.11802489227718777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,10240,0.1772853268517388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,51200,0.33971820937262637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,8192,0.14505333370632595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,10240,0.1045573353767395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,8192,0.06294222010506524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,7168,0.12712977992163763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,12288,0.21000444889068604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,7168,0.07400621970494588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,6144,0.044906665881474815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,12288,0.0806293355094062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,5120,0.09305066532558864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,6144,0.06306666798061795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,10240,0.06981333096822102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,5120,0.05306577682495117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,5120,0.03856799999872843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,3584,0.06720177994834052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,4096,0.03285955720477634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,4096,0.0440924432542589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,3072,0.059005333317650684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,3584,0.029818667305840388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,3584,0.039664000272750854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,2560,0.049545778168572314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,3072,0.02688088847531213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,3072,0.03470755616823832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,2048,0.040983110666275024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,2560,0.02346311178472307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,4096,0.07576266924540202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,1536,0.0326257778538598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,2048,0.02293955617480808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,2048,0.026378666361172993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,7168,0.060179558065202504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,1024,0.024176888995700415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,1536,0.033279998434914485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,1024,0.014718222949239941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,768,0.017815111411942374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,51200,0.47275378968980575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,2560,0.030125333203209773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,512,0.012896888785892062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,6144,0.1090186635653178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,256,0.008835555778609382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,512,0.013793778088357715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,1536,0.016566221912701923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,128,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,256,0.010749333434634738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,128,0.010042666561073726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,128,0.010094222095277574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,64,0.005214222189452913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,768,0.012308444413873883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,8192,0.08503911230299209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,512,0.011031111081441244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,64,0.010020444790522257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,32,0.013140444126394061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,256,0.010026666853162978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,51200,0.2612053288353814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,65536,0.22507998678419325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,65536,0.1502995491027832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,768,0.018135999639829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,32,0.0052933332820733385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,51200,0.11479732725355361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,16384,0.041048887703153826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,65536,0.3337635464138455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,12288,0.032510221004486084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,12288,0.045604444212383695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,16384,0.08686311377419366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,8192,0.047028445535235934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,12288,0.06738133562935723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,16384,0.06040444638994005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,8192,0.034040000703599714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,10240,0.055431998438305326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,7168,0.0296053323480818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,6144,0.03525422347916497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,10240,0.027839110957251653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,1024,0.024039111203617517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,6144,0.019480889042218525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,5120,0.03070577647950914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,7168,0.04082489013671875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,5120,0.017652443713612027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,5120,0.022827555735905964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,4096,0.025751999682850305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,7168,0.024259555670950148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,4096,0.015841777125994366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,51200,0.17263021734025744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,3584,0.014519110321998596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,3584,0.01723911033736335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,6144,0.02644799980852339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,8192,0.024784889486100938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,2560,0.01687022215790219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,3072,0.01309866706530253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,3072,0.015517334143320719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,2048,0.014105778601434497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,2560,0.012150222228633033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,2560,0.013769778112570444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,10240,0.03948622279696994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,3584,0.022614222433831956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,4096,0.01898755629857381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,2048,0.012442666623327466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,1536,0.014008000493049622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,1024,0.008373333348168267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,3072,0.019907555646366544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,1024,0.011405333048767514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,512,0.005961777849329843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,768,0.007727999654081132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,768,0.009399111072222391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,256,0.004290666845109728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,512,0.008012444608741337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,1536,0.011408888631396823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,2048,0.01223288890388277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,256,0.006673777682913675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,1536,0.009451555709044138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,64,0.0033093332830402586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,128,0.0070222218831380205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,768,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,64,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,32,0.007874666816658443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,512,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,65536,0.2725377877553304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,128,0.0039022221333450745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,51200,0.212409774462382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,65536,0.17358044783274332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,16384,0.0711724427011278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,128,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,51200,0.13201155927446154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,12288,0.05594399902555677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,16384,0.03799110982153151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,16384,0.04729511009322273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,10240,0.04607999987072415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,12288,0.02936977810329861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,12288,0.036152889331181846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,8192,0.0389120015833113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,256,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,10240,0.026757333013746474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,7168,0.03393066591686673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,8192,0.023362666368484497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,8192,0.027449778384632532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,32,0.0035786665976047516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,6144,0.029489778810077246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,7168,0.022672888305452134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,7168,0.023870221442646448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,51200,0.10412089029947917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,6144,0.018215111560291715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,4096,0.022301332818137273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,5120,0.016188444362746347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,5120,0.017991110682487488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,3584,0.018920888503392536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,4096,0.014409777190950183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,4096,0.015587555037604438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,3072,0.01687466601530711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,65536,0.12894311216142443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,3584,0.013343999783198038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,3584,0.014134221606784396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,3072,0.013425778183672162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,3072,0.01310488912794325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,5120,0.025822222232818604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,2560,0.011562666959232755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,2560,0.01202311118443807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,1536,0.01034311122364468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,2048,0.010579555398888059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,2048,0.01073688848151101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,1024,0.00868888861603207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,1536,0.01220533334546619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,10240,0.031311111317740545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,1024,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,2560,0.014497778481907315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,768,0.007673777639865875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,2048,0.012434666355450949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,512,0.005547555370463266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,6144,0.021271111236678228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,512,0.007291555404663086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,512,0.006646222124497096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,1536,0.00868000007337994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,128,0.0038737778862317405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,256,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,256,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,768,0.007291555404663086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,1024,0.009366222553782994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,64,0.0032479998966058097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,128,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,32,0.003244444520937072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,768,0.00817599975400501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,64,0.006002666635645761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,32,0.007299555672539606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,65536,0.24993867344326445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,51200,0.16153866714901394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,65536,0.1647919946246677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,16384,0.05771199862162272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,51200,0.12671111689673528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,128,0.006685333533419504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,12288,0.0444088876247406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,16384,0.035523557000690036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,12288,0.028246220615175035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,256,0.0041413332025210066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,10240,0.03885511226124234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,65536,0.12177777290344238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,10240,0.02457688914404975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,10240,0.030402666992611353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,51200,0.09800533453623454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,8192,0.021324444148275588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,7168,0.027781334188249376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,16384,0.0450524455971188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,6144,0.02499288817246755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,7168,0.019288000133302476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,7168,0.022642667094866436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,12288,0.03467377689149644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,6144,0.017589333984586928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,4096,0.017654221918847825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,6144,0.01999200052685208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,5120,0.015726221932305228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,1024,0.00925333301226298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,3584,0.016155555844306946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,5120,0.017267554998397827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,8192,0.03188088867399428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,3072,0.014463111758232117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,4096,0.015056000815497505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,2560,0.012471111284361945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,3072,0.01219733307758967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,8192,0.025808889004919264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,3072,0.012749333348539142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,2048,0.011386666860845355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,2560,0.010596444209416708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,1536,0.009369778136412302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,2560,0.011557333171367645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,2048,0.00980088859796524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,5120,0.020979555116759405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,2048,0.010602666272057427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,1536,0.008557333714432186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,768,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,1536,0.011051555474599203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,1024,0.007496889266702864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,512,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,1024,0.00905511114332411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,768,0.007153777612580194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,256,0.004211555752489301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,768,0.007646222081449296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,512,0.007068444457319047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,3584,0.013918222652541267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,512,0.006778666542636023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,256,0.006801777829726537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,64,0.0031911111954185697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,256,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,128,0.006611555400821898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,32,0.0032071111102898917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,1024,0.007679999702506595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,128,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,65536,0.16983021630181205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,64,0.006116444451941385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,32,0.006612444503439798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,51200,0.13424711757236057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,4096,0.01384888920519087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,65536,0.12285155720180935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,3584,0.013999111122555204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,51200,0.057088891665140785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,12288,0.03717333409521315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,16384,0.023357333408461675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,16384,0.03384622269206577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,10240,0.031164444155163232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,12288,0.0188080006175571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,12288,0.02683111031850179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,8192,0.026368889543745253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,10240,0.023352889551056757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,8192,0.020264888803164165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,65536,0.07073244121339586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,51200,0.09265689055124919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,7168,0.017455110947291057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,6144,0.020311110549502902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,6144,0.012759110993809171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,6144,0.015855110353893705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,10240,0.016536888149049547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,7168,0.02327999969323476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,5120,0.013933334085676404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,4096,0.015526221858130561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,16384,0.047898666726218335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,3584,0.013439999686347114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,4096,0.01070666644308302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,7168,0.013669333524174161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,3584,0.009416888985368941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,3584,0.011630222201347351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,3072,0.012057777908113269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,5120,0.017475555340449016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,2560,0.010512000156773461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,3072,0.009030222064918941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,8192,0.014907555447684394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,2048,0.009335999687512716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,2560,0.010012444522645738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,5120,0.01165422217713462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,1536,0.007831999825106727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,2048,0.008336000144481659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,2048,0.008783111141787635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,1024,0.0063866666621632045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,1536,0.007324444750944774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,1536,0.009342222577995723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,1024,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,1024,0.007677333222495184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,3072,0.010727110836240979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,768,0.006587555425034628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,768,0.0069573331210348345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,256,0.003519111209445529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,512,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,512,0.006126222097211414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,256,0.006108444597986009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,256,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,128,0.005975111077229182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,768,0.0047084444926844705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,128,0.00544177782204416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,32,0.003543111185232798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,512,0.0038853333228164246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,64,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,32,0.006002666635645761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,128,0.0035422220826148987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,128,0.003540444291300244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,65536,0.06809066401587592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,51200,0.12182488706376816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,16384,0.043576889567905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,64,0.0032213332338465583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,51200,0.0550764434867435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,12288,0.0346684455871582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,16384,0.022260444031821355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,2560,0.008226667013433244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,16384,0.033478223615222506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,10240,0.0288355549176534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,12288,0.018000889155599806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,65536,0.15466666221618652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,12288,0.025834666358100042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,10240,0.016050666570663452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,4096,0.012678222523795234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,7168,0.02166755497455597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,8192,0.01423733267519209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,65536,0.11815199587080215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,8192,0.01959111127588484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,7168,0.017133333616786532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,6144,0.012359999948077731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,6144,0.01571644511487749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,5120,0.011408888631396823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,4096,0.01424533294306861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,8192,0.02462133268515269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,10240,0.02240444388654497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,4096,0.009863999982674917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,3584,0.012688889271683164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,6144,0.019003555178642273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,4096,0.012479110724396177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,7168,0.013170666992664337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,3584,0.009387555221716562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,3584,0.011395555403497485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,3072,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,3072,0.01075644459989336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,2048,0.008773333496517604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,2560,0.009679111341635386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,2048,0.007635555333561367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,2048,0.008396444221337637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,3072,0.011331555744012197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,1536,0.00908088849650489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,2560,0.010131555298964182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,2560,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,5120,0.013795555465751223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,768,0.0052471111218134565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,1536,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,768,0.006802666518423293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,512,0.0042168887125121225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,5120,0.01646133263905843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,512,0.006086222413513396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,256,0.0035377778112888336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,1024,0.007694222033023834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,1024,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,128,0.003192000091075897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,256,0.006001777946949005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,256,0.005586666779385672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,64,0.0031893334041039148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,128,0.005985777825117111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,768,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,128,0.005617777920431561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,64,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,65536,0.14112888442145452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,32,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,512,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,65536,0.06576710939407349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,65536,0.11247732904222275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,51200,0.05327999922964308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,51200,0.0862942205535041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,16384,0.04215022259288364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,51200,0.09109244743982951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,16384,0.02124888863828447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,12288,0.033448888195885554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,12288,0.017515555024147034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,51200,0.11296889517042373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,10240,0.029224889145957098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,10240,0.015658666690190632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,8192,0.023039999935362074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,10240,0.021321778496106465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,8192,0.013771555489963956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,7168,0.021846221552954778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,16384,0.0320355561044481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,7168,0.012920888761679331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,12288,0.024655999408827886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,7168,0.016773333152135212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,5120,0.015481778317027621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,6144,0.011984888878133563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,6144,0.015253333581818474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,32,0.0032337777730491427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,4096,0.012848888834317526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,5120,0.013464888764752282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,1024,0.005697777701748743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,4096,0.009347555538018545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,8192,0.018734221657117207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,4096,0.012121777567598553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,3072,0.01037599974208408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,3584,0.008623999853928884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,2560,0.00962488849957784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,3072,0.008288000192907121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,3072,0.010499555203649731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,2048,0.008642666869693333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,2560,0.007679999702506595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,2560,0.009001777403884465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,1536,0.007300444775157505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,2048,0.007491555478837755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,2048,0.00812977800766627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,1024,0.005469333380460739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,1536,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,1536,0.008452444440788692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,768,0.00490399988161193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,1024,0.006648889018429651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,1024,0.007332444190979004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,512,0.004223111189073986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,768,0.00658044425977601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,512,0.006132444573773279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,512,0.005997333261701796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,5120,0.010809777511490716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,1536,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,256,0.005963555640644497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,128,0.0032186667538351486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,256,0.005669333454635408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,64,0.003176888864901331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,128,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,3584,0.011305777562989129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,32,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,6144,0.017239999439981248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,64,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,32,0.006020444548792309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,256,0.0035502223504914176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,65536,0.06389066908094618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,51200,0.08843466970655654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,16384,0.034703999757766724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,3584,0.012078222301271228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,51200,0.05105511016315884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,51200,0.08377422226799859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,16384,0.02047644390000237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,16384,0.030943112240897283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,128,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,10240,0.025779555241266888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,12288,0.02363288899262746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,65536,0.11097688807381524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,8192,0.022239999638663396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,10240,0.01516711049609714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,7168,0.01983644399378035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,8192,0.013395555317401886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,65536,0.10858044359419082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,8192,0.018073777357737224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,7168,0.012413333687517377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,7168,0.0163128889269299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,6144,0.011388444238238864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,5120,0.012825777961148156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,6144,0.015099555253982544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,4096,0.011391999820868174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,5120,0.010042666561073726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,5120,0.013193777865833707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,3584,0.010383110907342698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,4096,0.009025777379671732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,4096,0.011838222543398539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,10240,0.020802666743596394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,3584,0.008370666868156856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,6144,0.014523555835088095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,3584,0.010888000329335531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,2560,0.00979644474056032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,3072,0.008035555481910706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,3072,0.009888889061080085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,2048,0.008301333420806462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,2560,0.00867377801073922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,768,0.006311111152172089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,2048,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,3072,0.011148444480366178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,1536,0.007991111112965478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,1024,0.0052355556852287715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,1536,0.0069306666652361555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,1024,0.006405333264006509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,1024,0.006993778049945831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,12288,0.016933333542611863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,1536,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,512,0.004360888981156879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,768,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,512,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,256,0.0035422220826148987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,12288,0.030237333642111883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,256,0.005607999861240387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,128,0.0032231110251612137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,256,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,64,0.002883555574549569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,128,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,2048,0.007663999994595845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,32,0.0029297777348094513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,768,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,64,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,65536,0.09239911370807224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,512,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,32,0.005967999911970562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,2560,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,65536,0.06123555368847317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,51200,0.05199022094408671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,16384,0.030894223186704848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,12288,0.02472177810139126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,16384,0.019840000404251944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,128,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,16384,0.02019999921321869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,12288,0.016303999556435477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,12288,0.01611288885275523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,8192,0.015764445066452026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,10240,0.014627556006113688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,10240,0.014527999692493014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,51200,0.07286577754550509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,8192,0.012763555679056378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,8192,0.013287111288971372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,768,0.004563555535342959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,6144,0.01276711126168569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,7168,0.011773332953453064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,5120,0.011397333608733283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,6144,0.010709332923094431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,6144,0.011144888897736868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,4096,0.010392000277837118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,5120,0.00997866690158844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,7168,0.011797332929240333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,3584,0.009008888569143083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,4096,0.009020444419648912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,65536,0.06775199704700045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,7168,0.014287999934620328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,3072,0.008661333057615492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,3584,0.008096000386608971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,2560,0.007848000360859765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,3584,0.008375999828179678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,3072,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,2048,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,3072,0.007423111134105259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,2560,0.007668444679843054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,1536,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,10240,0.02176977859603034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,2048,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,1024,0.004234666625658671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,1536,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,4096,0.008630221916569604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,1536,0.0070053330726093715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,768,0.0038942222793896994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,1024,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,512,0.003551111039188173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,1024,0.005993777679072485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,768,0.006242666807439592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,256,0.003208000005947219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,768,0.005669333454635408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,512,0.005990222096443176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,128,0.0030942221896515954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,512,0.0052977779673205475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,256,0.005283555636803309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,128,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,64,0.0031635556370019913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,128,0.005193777796294955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,64,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,65536,0.07916800181070964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,51200,0.048913776874542236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,256,0.005971555494599872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,65536,0.05998400184843275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,2560,0.006745777610275481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,5120,0.00959288908375634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,16384,0.028488000233968098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,51200,0.048157334327697754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,51200,0.05022755596372816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,16384,0.019653333557976615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,32,0.005628444254398346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,16384,0.01923111081123352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,12288,0.016172443826993305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,12288,0.015596444408098856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,10240,0.02014666630162133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,65536,0.06475644641452365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,8192,0.018031110366185505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,51200,0.06392266352971394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,10240,0.014077333940400017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,32,0.0031902222997612427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,7168,0.013781333135233985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,8192,0.01274400038851632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,6144,0.012364444633324942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,12288,0.023081777824295893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,7168,0.011397333608733283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,5120,0.011065777805116443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,6144,0.010081777969996134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,6144,0.010692444112565784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,4096,0.010028444230556488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,5120,0.009686222506894005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,10240,0.01443377799457974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,3584,0.00871288859181934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,4096,0.008351999852392409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,8192,0.012469333079126147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,3072,0.008368000388145447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,3584,0.00815288888083564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,7168,0.011078221930397881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,3584,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,3072,0.0076657773719893555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,2048,0.006267555471923616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,2560,0.00684799998998642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,5120,0.009184888667530483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,1536,0.00498488876554701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,2048,0.007303111255168915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,1536,0.006641777853171031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,1024,0.004454222404294544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,2560,0.007448888487286038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,1024,0.005965333431959152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,3072,0.007335111498832703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,1024,0.006559999866618051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,768,0.006267555471923616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,512,0.003583999971548716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,2048,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,4096,0.00832088871134652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,1536,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,256,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,2560,0.007617777420414819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,128,0.0031902222997612427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,128,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,128,0.00592533333433999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,64,0.0031724443866146933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,768,0.003990222182538775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,64,0.0052471111218134565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,32,0.00291377781993813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,65536,0.06912000311745538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,32,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,65536,0.061956445376078285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,65536,0.05929333633846707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,51200,0.05580088827345106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,51200,0.046910223033693105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,256,0.00590844452381134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,16384,0.0185217774576611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,16384,0.01918577816751268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,12288,0.02034311162100898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,768,0.005653333332803514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,10240,0.01724444495307075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,12288,0.015878222054905362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,10240,0.014133334159851074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,8192,0.015184889237085978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,10240,0.013531555732091268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,8192,0.012117333710193634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,7168,0.013329777452680798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,256,0.003544888976547453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,7168,0.011146666275130378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,6144,0.011821333732869891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,7168,0.010889777706729041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,6144,0.010020444790522257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,5120,0.01075466639465756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,8192,0.012381333443853589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,5120,0.009359999663299983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,4096,0.009687110781669617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,5120,0.009008888569143083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,512,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,3584,0.008643555144468943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,4096,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,4096,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,3072,0.0081386665503184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,3584,0.008055999875068665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,3584,0.007646222081449296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,6144,0.010567111273606619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,3072,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,3072,0.007296000089910295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,512,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,2560,0.006656888872385025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,51200,0.04814489020241631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,2048,0.0063484443558586975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,1536,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,2560,0.007282666862010956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,1024,0.004527111020353106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,1536,0.006635555376609166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,2560,0.007407111426194509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,1024,0.006343110981914733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,1024,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,768,0.0038924444880750445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,2048,0.007334222396214803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,1536,0.005473777651786804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,512,0.003556444413132138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,768,0.006303999986913469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,256,0.003185777821474605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,512,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,512,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,256,0.005969777703285217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,256,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,2048,0.006609777609507243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,128,0.0058248887459437055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,128,0.004979555391603046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,32,0.0028897778441508612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,12288,0.015062222878138224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,64,0.005245333330498801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,32,0.0052844443255000645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,768,0.005633777628342311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,65536,0.059262222713894315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,64,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,51200,0.04633955491913689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,16384,0.021984888447655573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,51200,0.045765333705478244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,65536,0.0648897753821479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,16384,0.018992000155978732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,16384,0.01811911165714264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,10240,0.015655999382336933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,12288,0.014840000205569796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,10240,0.013453333742088742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,8192,0.014096000128322177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,16384,0.02402577797571818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,8192,0.011389333340856763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,8192,0.011983999775515662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,7168,0.01347644461525811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,12288,0.015544000599119397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,6144,0.011397333608733283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,51200,0.05380533470047844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,10240,0.013377777404255338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,6144,0.01020088874631458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,5120,0.01012711144155926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,5120,0.009008000294367472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,4096,0.009189333352777693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,12288,0.018253333038753934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,4096,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,4096,0.007805333369308048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,128,0.002842666581273079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,3584,0.008055999875068665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,6144,0.00979644474056032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,5120,0.008716444174448649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,3072,0.008015111088752747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,2560,0.0069511110583941145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,3584,0.008336000144481659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,65536,0.05817066960864597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,2048,0.005923555543025334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,2560,0.007354666789372762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,2560,0.006611555400821898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,3072,0.007979555262459649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,2048,0.007066667079925537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,2048,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,1024,0.0042088888585567474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,1536,0.006065777606434292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,1024,0.00657155571712388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,768,0.0038728887836138406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,7168,0.010724444356229572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,768,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,768,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,1536,0.004907555464241239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,512,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,256,0.0031973332580592898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,512,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,256,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,128,0.003034666594531801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,7168,0.011094222466150919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,128,0.004935111022657818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,3584,0.007387555307812161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,64,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,32,0.0028417776856157514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,1536,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,64,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,65536,0.05323644479115804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,32,0.0052426668504873914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,51200,0.043303112188975014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,1024,0.0059164443777667145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,65536,0.056482666068606906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,16384,0.01787644459141625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,51200,0.04509244362513224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,51200,0.044504887527889676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,128,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,512,0.003552000141806073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,16384,0.018617777360810172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,16384,0.017616000440385606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,10240,0.014135110709402295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,12288,0.014430221584108142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,8192,0.013456000222100152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,12288,0.014530667000346713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,10240,0.01314399970902337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,8192,0.011115555961926779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,7168,0.01204533295498954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,256,0.004907555464241239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,65536,0.056242664655049644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,6144,0.011345778074529437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,7168,0.01037599974208408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,7168,0.010375111467308467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,12288,0.014821334017647637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,6144,0.010025777750545079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,6144,0.009420444567998251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,4096,0.00868088917599784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,5120,0.008554666406578487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,10240,0.012498666842778524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,3584,0.007978666987684038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,4096,0.00833422193924586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,8192,0.011414222419261932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,3584,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,3584,0.007384888827800751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,5120,0.010077333284748925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,3072,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,5120,0.00907199995385276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,3072,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,2560,0.007440889047251807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,4096,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,1536,0.00490399988161193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,3072,0.007062222394678328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,2048,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,1024,0.00424799985355801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,1536,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,1536,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,3072,0.008021333151393467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,2560,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,1024,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,512,0.003540444291300244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,768,0.006143110907740063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,768,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,2048,0.005583111196756363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,512,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,2048,0.007244444555706448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,128,0.0028488888508743713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,256,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,256,0.004960888789759742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,64,0.0029093333416514923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,128,0.005767999837795894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,32,0.0028702221396896574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,64,0.0052844443255000645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,65536,0.04193333453602261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,2560,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,65536,0.0554328891966078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,65536,0.054452445771959096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,51200,0.03358577688535055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,51200,0.04474844535191854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,128,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,16384,0.01516888870133294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,1024,0.005621333503060871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,12288,0.014734221829308404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,32,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,16384,0.01748533381356133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,256,0.0032231110251612137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,12288,0.013603555659453074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,12288,0.01425244410832723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,8192,0.012444444828563266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,10240,0.012474666866991254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,768,0.003896000070704354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,8192,0.010394666757848527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,16384,0.016865778300497267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,7168,0.010050666828950247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,512,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,7168,0.010391111175219217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,6144,0.009686222506894005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,5120,0.010383110907342698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,8192,0.011114666859308878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,51200,0.04288355509440104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,5120,0.008353778057628209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,5120,0.009082666701740688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,4096,0.008479110896587372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,4096,0.007654222349325816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,3584,0.007337777978844113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,3584,0.008399999803966945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,3072,0.006528888725572162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,7168,0.011400889191362592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,3072,0.0070159998204973005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,2560,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,3072,0.007718222008811102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,2560,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,2048,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,10240,0.012422222230169507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,6144,0.009362666971153682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,2048,0.0062977779242727495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,2048,0.0070773329999711775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,1536,0.006676444576846228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,4096,0.009127111070685917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,1024,0.003998222036494149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,1024,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,768,0.003814222084151374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,1024,0.005655111124118169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,768,0.006304889089531369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,512,0.003237333355678452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,1536,0.004593777573770947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,512,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,512,0.0052897776994440295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,1536,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,10240,0.013447110851605734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,256,0.0052826665341854095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,128,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,256,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,3584,0.007354666789372762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,128,0.0059164443777667145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,64,0.002524444419476721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,32,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,256,0.0032257777121331957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,64,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,65536,0.030829333596759375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,32,0.005028444445795483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,51200,0.025220443805058796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,6144,0.01074044406414032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,65536,0.051832887861463756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,16384,0.012847999731699625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,51200,0.044032888280020825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,51200,0.04110666778352525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,768,0.005587555468082428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,12288,0.010839111275143094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,16384,0.01586577792962392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,10240,0.01108266661564509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,12288,0.013462222284740873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,10240,0.012187555432319641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,10240,0.011747555600272285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,65536,0.054528001281950206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,2560,0.006816000160243776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,7168,0.012770666844315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,8192,0.011092444260915121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,8192,0.010423999693658618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,6144,0.011077333655622272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,7168,0.010424888796276517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,5120,0.009744000103738572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,128,0.004958222309748332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,6144,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,4096,0.009714666340086196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,12288,0.0131217779384719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,5120,0.008387555678685507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,4096,0.00832622249921163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,8192,0.013272888958454132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,3584,0.007827555967701806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,3072,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,16384,0.01684888866212633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,3072,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,6144,0.009351111120647853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,5120,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,3584,0.007037333316273159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,4096,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,3584,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,2048,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,2048,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,2560,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,3072,0.007672000262472365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,1024,0.004397333496146732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,1536,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,2560,0.00741244438621733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,2048,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,1024,0.0063164445261160536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,1024,0.00563822231358952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,768,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,768,0.00544888898730278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,1536,0.005541333307822545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,512,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,256,0.0031999999450312722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,256,0.005664888769388199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,1536,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,256,0.005106666849719154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,2560,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,64,0.002879111096262932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,128,0.005592888842026393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,32,0.0032071111102898917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,512,0.0036186666952239144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,768,65536,0.02641155487961239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,64,0.005112888912359874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,32,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,768,51200,0.02153066628509098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,65536,0.05442666676309374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,65536,0.05086844497256809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,768,16384,0.011863999896579318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,51200,0.04381777842839559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,51200,0.04064088728692796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,768,0.00387999994887246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,768,12288,0.01072622256146537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,16384,0.016153777639071148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,128,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,12288,0.013456000222100152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,12288,0.013089777694808112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,768,8192,0.009364444348547194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,10240,0.012083555261294046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,10240,0.011632888681358762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,768,7168,0.00944977750380834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,8192,0.011021333436171213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,8192,0.010435555544164447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,768,6144,0.010083555347389644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,128,0.00291377781993813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,7168,0.009687999884287516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,768,5120,0.009005332986513773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,16384,0.0158924443854226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,6144,0.009081777599122789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,5120,0.009000000026490953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,768,4096,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,512,0.005976888868543837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,4096,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,768,3584,0.0069475554757648045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,7168,0.01018311083316803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,3584,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,768,3072,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,768,10240,0.009360888765917884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,5120,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,3072,0.00703022215101454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,7168,0.010092444717884064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,4096,0.008426666259765625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,2560,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,6144,0.00963199966483646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,3584,0.007287110719415877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,768,1536,0.004619555754794015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,2048,0.0069875551594628235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,3072,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,768,2560,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,1536,0.005982222242487802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,1024,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,768,2048,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,768,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,768,512,0.003569777641031477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,512,0.005296000176005893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,512,0.005980444451173146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,768,256,0.0032257777121331957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,1024,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,256,0.005892444401979446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,256,0.004938666605287128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,768,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,128,0.005848000033034219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,768,64,0.0025840000145965153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,128,0.004960000101062986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,768,1024,0.003951111187537511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,768,32,0.002562666725781229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,512,65536,0.02275111112329695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,64,0.004971555537647671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,32,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,768,768,0.0038888889054457345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,512,51200,0.01815555493036906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,65536,0.05358755588531494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,512,16384,0.012448000411192576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,768,128,0.0028497777465316984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,51200,0.04017688830693563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,512,12288,0.010046222143703038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,16384,0.016193777322769165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,16384,0.01552977826860216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,512,10240,0.01037599974208408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,1536,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,12288,0.012771555946932899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,512,8192,0.009391110804345872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,2560,0.007360888852013483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,10240,0.011395555403497485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,512,7168,0.009376000199053023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,8192,0.011071110765139261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,65536,0.050048000282711454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,8192,0.010400888820489248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,512,6144,0.007712889048788283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,7168,0.010399999717871347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,512,5120,0.008045333127180735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,51200,0.040170666244294904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,6144,0.009038222332795462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,5120,0.008974221845467886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,12288,0.013452444639470843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,5120,0.008358221915033128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,512,3584,0.0070062221752272705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,10240,0.012117333710193634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,4096,0.007641777396202087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,512,3072,0.006268444574541516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,3584,0.008001777860853408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,3584,0.0073724447025193115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,512,2560,0.005902222047249476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,2048,0.006325333482689328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,3072,0.0069511110583941145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,7168,0.009676444033781687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,512,2048,0.0052355556852287715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,2560,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,512,1536,0.004578666554556953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,2048,0.006372444331645966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,4096,0.008344000412358178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,1536,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,3072,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,1024,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,512,768,0.0035457776652442086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,1024,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,2560,0.006680888848172293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,768,0.005335111171007156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,1536,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,512,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,512,256,0.0032186667538351486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,6144,0.009720889230569204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,512,4096,0.007378666765160031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,512,512,0.003527111063400904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,512,1024,0.003930666794379552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,128,0.004958222309748332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,512,64,0.0025475554996066624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,256,0.00499199993080563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,512,32,0.0028657778683635923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,2048,0.006994666324721442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,512,128,0.0028479999552170434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,256,65536,0.019368888603316415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,32,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,512,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,65536,0.04707644383112589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,64,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,51200,0.03843555516666836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,768,0.005967999911970562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,256,16384,0.009703111317422655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,16384,0.016160888804329764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,256,12288,0.010027555955780877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,16384,0.015174221661355762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,128,0.005759999983840519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,12288,0.013461333182122974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,12288,0.01276088919904497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,256,8192,0.008443555898136562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,10240,0.012090666426552666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,10240,0.011400000088744693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,256,0.0058559998869895935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,8192,0.011044444309340583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,8192,0.010296888649463654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,256,6144,0.008632000121805403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,7168,0.01038577738735411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,7168,0.009695111049546136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,256,5120,0.008039111064540016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,6144,0.009684444301658207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,65536,0.04908355408244663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,256,10240,0.009349333743254343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,5120,0.009001777403884465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,5120,0.008360889222886827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,4096,0.008310221963458592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,4096,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,256,7168,0.008006222546100616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,51200,0.03784888982772827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,256,3072,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,3584,0.008004444340864817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,3584,0.007305777735180325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,3072,0.007650666766696506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,3072,0.006989333364698622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,256,2048,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,2560,0.007334222396214803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,256,51200,0.016579555140601266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,256,1536,0.004574222283230888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,2048,0.006984000404675801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,2048,0.006384000182151794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,256,1024,0.0038684445122877755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,1536,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,1536,0.0058400001790788435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,256,768,0.003549333247873518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,1024,0.006303999986913469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,256,2560,0.005630222045713001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,256,512,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,768,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,6144,0.009020444419648912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,256,4096,0.007603555917739868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,512,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,256,0.005664000080691443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,256,128,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,256,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,128,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,256,64,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,128,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,256,32,0.0028488888508743713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,2560,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,512,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,128,65536,0.015491555134455362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,256,3584,0.0069324444565508105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,128,51200,0.015201777219772339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,1024,0.005647111270162795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,51200,0.037464890215131975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,128,16384,0.00941066692272822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,51200,0.037695109844207764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,64,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,256,256,0.0028684443483750024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,16384,0.016184889607959323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,65536,0.046579556332694165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,16384,0.01552977826860216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,128,10240,0.008672888908121321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,65536,0.04638933473163181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,128,8192,0.008346666892369589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,10240,0.012105777859687805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,10240,0.011808888779746162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,768,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,8192,0.010031110710567897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,128,12288,0.009721777505344814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,7168,0.009521777431170145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,128,6144,0.008375999828179678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,6144,0.009662222531106737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,12288,0.01351466692156262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,6144,0.009015111459626092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,5120,0.009024000002278222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,128,4096,0.007339555356237624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,7168,0.010278222461541494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,4096,0.00832088871134652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,128,3584,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,12288,0.01278133359220293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,3584,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,128,3072,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,128,7168,0.009036444127559662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,3072,0.006959111326270633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,32,0.004906666775544484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,128,2560,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,3072,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,2560,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,128,2048,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,3584,0.007297777467303806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,8192,0.010724444356229572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,128,1536,0.004222222086456087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,2048,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,4096,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,1536,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,128,1024,0.003879111260175705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,1536,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,128,768,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,1024,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,128,5120,0.007829333345095316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,1024,0.005616888817813661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,2560,0.006687110910813014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,768,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,768,0.005296000176005893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,128,256,0.003113777687152227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,512,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,5120,0.008365333080291748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,512,0.00499288903342353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,2048,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,128,64,0.0028657778683635923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,128,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,128,32,0.0028773333049482773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,128,0.005024000174469418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,64,65536,0.016805332567956712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,64,0.00510044437315729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,64,51200,0.01516888870133294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,32,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,64,16384,0.009186666872766282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,65536,0.046441776884926685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,64,12288,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,51200,0.03719200028313531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,64,10240,0.009012444151772393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,256,0.005636444522274866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,64,8192,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,64,7168,0.007696888513035244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,10240,0.011402666568756104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,64,6144,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,8192,0.010037333601050908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,64,5120,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,128,512,0.0032177778581778207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,6144,0.009012444151772393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,64,4096,0.0063368889192740125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,5120,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,128,128,0.002564444517095884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,16384,0.015159999330838522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,4096,0.007650666766696506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,64,2560,0.0063902222447925145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,3584,0.007076444725195567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,64,2048,0.005655999812814925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,3072,0.0069919998447100324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,64,1536,0.004272888931963178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,2560,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,7168,0.009680888719028896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,64,1024,0.003608889049953885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,2048,0.0063386667105886675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,64,768,0.003536888708670934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,64,512,0.0032017777363459268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,1536,0.005614222337802251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,64,256,0.002862222285734283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,1024,0.005623999983072281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,768,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,64,3072,0.006664000037643645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,512,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,256,0.005249777601824866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,256,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,128,0.004621333132187526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,32,65536,0.016190222567982145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,64,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,32,51200,0.015154666370815702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,65536,0.04596622122658623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,32,16384,0.009015111459626092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,51200,0.03722933265897963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,32,12288,0.009351111120647853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,64,64,0.0028479999552170434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,16384,0.015174221661355762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,12288,0.012759110993809171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,32,8192,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,32,0.0046968890560997855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,10240,0.011398221883508893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,32,7168,0.007636444436179266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,32,6144,0.0070675553547011475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,8192,0.010047111246320937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,32,5120,0.006990222467316522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,64,3584,0.006302222195598815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,7168,0.0094035557574696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,32,4096,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,6144,0.008804444637563493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,5120,0.008106666306654612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,12288,0.012759110993809171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,4096,0.007616000043021307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,32,2560,0.005973333285914526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,3584,0.007137777904669444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,32,2048,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,3072,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,2560,0.0064177778032090925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,32,1536,0.004557333472702238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,2048,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,64,128,0.0028862222615215513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,32,3584,0.00599911105301645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,1536,0.005688000056478713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,32,768,0.003548444559176763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,32,512,0.0032311110860771606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,64,32,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,32,256,0.0029093333416514923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,768,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,32,128,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,512,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,256,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,128,0.0047048889100551605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,32,32,0.0026382222357723448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,64,0.004973333328962326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,32,3072,0.006292444550328785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,32,0.004687110996908612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,1024,0.005625777774386936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,32,64,0.0025315555847353404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,51200,1.1253955629136827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,51200,0.6004426744249132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,16384,0.36794222725762266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,32,10240,0.008649778034951951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,12288,0.2727031177944607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,16384,0.13425778018103704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,51200,0.3944008880191379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,16384,0.21084711286756727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,12288,0.1485191053814358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,10240,0.2168604400422838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,81,32,1024,0.003587555554178026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,10240,0.12552000416649714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,8192,0.17587822013431123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,10240,0.09228889147440593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,8192,0.1061591108640035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,8192,0.06979111168119642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,12288,0.2162968847486708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,7168,0.15254844559563532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,12288,0.10125422477722168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,10240,0.18137778176201713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,7168,0.09327022234598796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,8192,0.14728889200422499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,6144,0.07837244537141588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,6144,0.05340800020429823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,7168,0.12866844071282282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,16384,0.29047022925482857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,6144,0.11510133743286133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,5120,0.06698044141133626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,4096,0.09274222453435262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,5120,0.046444445848464966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,6144,0.13460800382826063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,7168,0.061463998423682324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,4096,0.054076443115870156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,4096,0.07916977670457628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,3584,0.049903111325369946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,5120,0.11102755864461263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,3072,0.06938666767544217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,3584,0.0350542234049903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,3584,0.07036978006362915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,2560,0.05850133630964491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,3072,0.04210933380656772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,3584,0.08072888851165771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,3072,0.06213511361016167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,2560,0.036666668123669095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,2048,0.047305779324637524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,51200,0.9337395562065972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,2560,0.027451554934183758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,2560,0.05407377746370104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,1536,0.0373342235883077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,2048,0.030237333642111883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,2048,0.031750222047170006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,2048,0.04664355516433716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,1536,0.039670222335391574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,5120,0.09409599834018284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,1024,0.029370665550231934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,768,0.021295999487241108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,1024,0.03477599885728624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,768,0.02207999924818675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,768,0.027268444498380024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,768,0.013620444469981723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,3072,0.0314702226055993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,4096,0.03941866755485535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,512,0.012648888760142855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,256,0.009372444616423713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,256,0.012443555725945367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,512,0.014816888504558139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,256,0.011049778097205691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,128,0.007682667010360294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,512,0.023285332653257582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,256,0.019301333361201815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,128,0.011410666836632622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,64,0.006710222197903528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,128,0.011345778074529437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,32,0.006657777975002925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,128,0.01595288846227858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,1536,0.02057422200838725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,64,0.011231111155615913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,32,0.015457777513398064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,512,0.0162426663769616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,1024,0.015435554915004306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,65536,0.624858644273546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,1024,0.028071999549865723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,51200,0.8538968827989367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,51200,0.4735111130608453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,65536,0.4175768958197699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,16384,0.2773004372914632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,1536,0.03937600056330363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,16384,0.16641155878702799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,51200,0.3151182333628337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,12288,0.209044443236457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,16384,0.10908177826139663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,12288,0.07912711302439372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,10240,0.17485777537027994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,65536,1.0923982196384006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,12288,0.12071732680002849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,8192,0.14166755146450466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,16384,0.24081778526306152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,10240,0.10195910930633545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,12288,0.1837937831878662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,8192,0.08546933200624253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,7168,0.12602044476403132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,8192,0.05623288949330648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,10240,0.1495306624306573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,6144,0.10860000054041545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,8192,0.12290578418307835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,7168,0.10960888862609863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,6144,0.06286844280030993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,5120,0.09148533476723565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,10240,0.07377333111233182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,5120,0.05313511027230156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,7168,0.07233066691292657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,51200,0.7253164185418023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,4096,0.07423466444015503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,65536,0.9602088928222656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,5120,0.037296000454160906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,3584,0.06599733564588758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,4096,0.04649244414435493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,6144,0.04483733243412442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,3584,0.03956266575389438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,3584,0.029613332615958318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,3072,0.05780888928307427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,3584,0.06055821975072225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,7168,0.05717955695258247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,2560,0.0488319993019104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,3072,0.034910221894582115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,6144,0.09355199999279445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,3072,0.05349066522386339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,2560,0.023564444647894964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,2560,0.030194666650560167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,4096,0.06759822368621826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,2048,0.02624977793958452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,1536,0.03186577889654372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,2048,0.04065333472357856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,5120,0.07971288760503133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,1536,0.03191377719243368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,1536,0.01791111131509145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,4096,0.03808888792991638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,2560,0.046550220913357206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,1024,0.023524444964196947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,1024,0.013824889229403602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,1024,0.029323554701275293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,3072,0.0255386663807763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,768,0.023394667439990576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,768,0.012299555871221753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,2048,0.03980799847178989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,512,0.014338667194048563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,2048,0.020799110333124798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,512,0.02015555567211575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,256,0.008344000412358178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,768,0.018266666266653273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,256,0.01072355525361167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,128,0.005462222215202119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,256,0.010043555663691627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,128,0.010047111246320937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,128,0.01426577733622657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,128,0.010037333601050908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,512,0.011156444748242697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,32,0.004649777793222004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,64,0.009995555712117089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,256,0.01686044368478987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,1024,0.023402666052182514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,32,0.01332266628742218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,65536,0.3222373326619466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,65536,0.22439199023776582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,768,0.01720977822939555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,65536,0.14123466279771593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,51200,0.17202488581339517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,1536,0.03411022159788344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,51200,0.11050133572684394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,16384,0.06047022342681885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,64,0.004558222161398994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,12288,0.06449244419733684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,16384,0.058839109208848744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,16384,0.041668444871902466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,51200,0.15545688735114202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,12288,0.03265688816706339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,512,0.012345777617560493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,12288,0.043374223841561206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,10240,0.03939288854598999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,10240,0.03790666659673055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,51200,0.25006400214301217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,8192,0.047107554144329496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,65536,0.19968444771236846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,8192,0.034322665797339544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,8192,0.02453777856296963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,7168,0.039408889081743025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,8192,0.034168889125188194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,7168,0.029451555675930444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,7168,0.021652445197105408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,6144,0.03470844361517165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,7168,0.02915999955601162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,10240,0.029559112257427637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,6144,0.01941866676012675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,6144,0.026383999321195815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,6144,0.0265600002474255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,10240,0.05407377746370104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,16384,0.083405335744222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,5120,0.017631111873520743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,5120,0.0232933329211341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,4096,0.015503111812803479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,4096,0.019003555178642273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,4096,0.020413332515292697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,5120,0.02952622373898824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,3584,0.01720355616675483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,3072,0.019256000717480976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,3584,0.015600888265503777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,5120,0.022767111659049988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,2560,0.01623644431432088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,3072,0.013726222018400827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,3072,0.017447999782032438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,3584,0.02201066745652093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,2560,0.012112889024946423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,2048,0.013468444347381592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,2560,0.015791111522250705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,2048,0.011974222130245633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,1536,0.011090666883521609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,3584,0.01905955539809333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,12288,0.045659555329216846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,3072,0.01551644504070282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,1536,0.013820444544156393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,1536,0.013468444347381592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,2560,0.013778666655222574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,1024,0.01145244472556644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,768,0.0073520003093613525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,1024,0.011891555454995898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,2048,0.012483555409643384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,2048,0.014460444450378418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,512,0.006114666660626729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,768,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,768,0.010402667025725046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,1024,0.008650666309727563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,256,0.0038924444880750445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,512,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,512,0.009388444324334463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,256,0.006950221955776215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,1024,0.008055999875068665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,256,0.008047111332416534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,128,0.006314666734801398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,4096,0.024845333562956914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,64,0.0032462223122517266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,128,0.008011555506123437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,128,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,32,0.003491555651028951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,512,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,64,0.006664000037643645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,65536,0.25344355901082355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,256,0.006647999915811751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,128,0.003576888806290097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,65536,0.1723386711544461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,51200,0.19782311386532256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,768,0.010048888623714447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,65536,0.12893333699968126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,51200,0.13144000371297201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,16384,0.07053422265582614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,1536,0.00963288876745436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,51200,0.1027911106745402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,12288,0.0558791094356113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,16384,0.037232001622517906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,16384,0.047545777426825635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,32,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,16384,0.04860799842410617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,12288,0.03829777903027005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,10240,0.03125244379043579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,10240,0.025794666674402025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,10240,0.033441778686311506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,8192,0.027779555983013574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,8192,0.029366221692827012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,10240,0.04619733492533366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,7168,0.034027556578318276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,12288,0.029614223374260798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,51200,0.12602844503190783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,8192,0.03933599922392104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,6144,0.029645333687464397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,7168,0.02128533356719547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,6144,0.018156444032986958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,5120,0.025628444221284654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,8192,0.021977777282396953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,12288,0.036293334431118436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,65536,0.16433244281344944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,5120,0.020896888441509668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,5120,0.016427555018001132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,7168,0.023864888482623633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,6144,0.02107288936773936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,4096,0.0155395550860299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,6144,0.02352977792421977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,4096,0.014465777410401238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,5120,0.018283555905024212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,3584,0.01722577710946401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,3584,0.014187556174066333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,3072,0.016538666354285348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,7168,0.026015111141734656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,3072,0.01240533341964086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,3072,0.013092444174819522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,2560,0.013997332917319404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,3072,0.01591822173860338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,4096,0.018555555078718398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,2048,0.012145777543385824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,2560,0.01165600038237042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,2560,0.011843555503421359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,3584,0.014182221558358936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,2048,0.010384000009960597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,2048,0.010840000377760993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,2048,0.013781333135233985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,1536,0.012457778056462606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,1536,0.012780444489585029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,3584,0.019254222512245178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,1024,0.009489778015348647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,1024,0.008050666915045844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,768,0.0069413334131240845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,1024,0.011122666299343109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,2560,0.015236443943447538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,512,0.005558222118351195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,768,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,768,0.008374222450786168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,768,0.009711110757456886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,256,0.0039013334446483185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,1536,0.00905511114332411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,512,0.00851288851764467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,256,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,128,0.0035546666218174826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,256,0.0083155557513237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,1536,0.01034222212102678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,128,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,64,0.003576000117593341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,128,0.006664888726340399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,32,0.003536888708670934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,512,0.007119110888904995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,512,0.007342221836249034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,1024,0.008379555410808986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,65536,0.2152142259809706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,256,0.006707555717892117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,65536,0.16448267300923666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,51200,0.17007111178504097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,65536,0.12041332986619736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,4096,0.022425777382320825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,16384,0.05678133169809977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,51200,0.09638133313920762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,65536,0.1478382216559516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,32,0.007078222102589077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,12288,0.04403466648525662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,51200,0.1175795528623793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,16384,0.04604355494181315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,12288,0.03457066747877333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,12288,0.028024001253975764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,12288,0.03620533479584588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,64,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,10240,0.030236442883809406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,8192,0.03185955683390299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,10240,0.03171199891302321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,16384,0.035564445787005954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,8192,0.025992888543340895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,8192,0.02108088963561588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,128,0.008232889076073965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,51200,0.12668532795376247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,7168,0.022600000103314716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,7168,0.019640889432695176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,6144,0.02539288832081689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,7168,0.025012445118692186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,16384,0.04485955503251818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,10240,0.024326221810446844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,5120,0.021389333738221064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,6144,0.017496888836224873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,6144,0.019908444748984445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,4096,0.018333334061834548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,5120,0.015822221835454304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,5120,0.017304889029926725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,7168,0.029373334513770208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,5120,0.02051111062367757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,10240,0.03748977846569485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,4096,0.014007111390431723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,4096,0.015125332607163323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,4096,0.017986666825082567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,3584,0.013441777891582914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,3584,0.016933333542611863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,3072,0.012803555362754397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,3072,0.012080888781282636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,2560,0.012818666795889536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,3072,0.01585155559910668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,6144,0.022597332795461018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,8192,0.02757511039574941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,2560,0.011727111207114326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,3584,0.016209777858522203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,2560,0.014455111490355598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,3072,0.014394667413499622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,3584,0.013891556196742587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,2048,0.01385600037044949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,1536,0.01071733319097095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,1536,0.012626666989591388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,1536,0.00832088871134652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,2560,0.010962666736708747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,2048,0.011408000356621213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,1024,0.00741244438621733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,768,0.006520888871616787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,1024,0.01053066634469562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,768,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,768,0.007641777396202087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,512,0.0052933332820733385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,1024,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,512,0.006685333533419504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,512,0.006993778049945831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,512,0.009031111167536842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,1024,0.008904889225959778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,1536,0.009004444711738164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,128,0.003569777641031477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,256,0.006320000108745363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,256,0.008104888929261101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,128,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,64,0.0032213332338465583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,128,0.008372444245550368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,32,0.0032142222755485107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,256,0.004253333227501975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,64,0.006129777679840724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,2048,0.009602666729026372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,32,0.007024889190991719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,2048,0.010659555594126383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,65536,0.17731644047631157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,256,0.006729777902364731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,51200,0.13440177175733778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,65536,0.12007999420166016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,65536,0.07101422548294067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,16384,0.047071999973720975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,128,0.005980444451173146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,51200,0.057295110490587026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,12288,0.03760622276200189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,16384,0.03427911135885451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,16384,0.023152889476882085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,16384,0.04288444585270352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,65536,0.13690488868289524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,12288,0.018792889184421964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,12288,0.02680355476008521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,12288,0.0340666671593984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,10240,0.01651022169325087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,10240,0.030037333567937214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,768,0.009703111317422655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,51200,0.09246400329801772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,7168,0.023408000667889912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,8192,0.014525334040323893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,8192,0.025993777645958796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,51200,0.10904266436894734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,7168,0.017550221747822232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,6144,0.020615110794703167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,7168,0.023338667220539514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,10240,0.023063111636373732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,5120,0.01756977703836229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,6144,0.021547555923461914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,8192,0.020058666666348774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,5120,0.013787555197874704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,4096,0.015239111251301236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,5120,0.011666666302416058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,10240,0.031239112218221027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,7168,0.013591110706329346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,3584,0.013282666603724161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,4096,0.017268444101015728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,8192,0.026517333255873785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,6144,0.015912888778580558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,3072,0.01220533334546619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,3584,0.011778666741318174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,3584,0.01604977746804555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,3072,0.010712888505723743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,5120,0.019297776950730216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,3072,0.009298666483826106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,4096,0.010831111007266574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,2560,0.00981777740849389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,2560,0.014132445057233175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,2560,0.008532444636027018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,3584,0.009398221969604492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,2048,0.009360888765917884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,1536,0.007613333563009898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,2048,0.01313688854376475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,2560,0.010612444745169746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,1536,0.00907022257645925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,1536,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,4096,0.012659555508030785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,2048,0.009522666533788046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,6144,0.012607999973826937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,768,0.005223999834722943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,1024,0.009672889278994666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,2048,0.008009778128729926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,768,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,3072,0.015019555886586508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,768,0.00903200027015474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,512,0.004253333227501975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,768,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,1024,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,1536,0.011395555403497485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,512,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,512,0.006413333531883027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,512,0.008716444174448649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,128,0.0032319999817344877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,1024,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,256,0.008035555481910706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,128,0.005994666781690385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,128,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,256,0.003555555724435382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,64,0.005688888745175467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,32,0.0032008888406885993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,65536,0.1533448828591241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,256,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,32,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,128,0.00562755556570159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,65536,0.11712355083889431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,64,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,51200,0.12092533376481797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,65536,0.0685031082895067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,16384,0.047263112333085805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,51200,0.05507733424504598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,65536,0.1363919973373413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,1024,0.007663999994595845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,12288,0.03511200017399258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,51200,0.1082000003920661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,16384,0.04189244574970669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,10240,0.029183109601338703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,12288,0.025773333178626165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,12288,0.03338222371207343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,10240,0.022299556268586054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,8192,0.024733333124054804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,51200,0.09102221992280747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,256,0.005790222022268508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,8192,0.019578667150603402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,7168,0.02177511155605316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,8192,0.02557777696185642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,12288,0.018089777893490262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,7168,0.013230221966902414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,6144,0.019298666053348117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,10240,0.029370665550231934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,6144,0.015526221858130561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,16384,0.03285600079430474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,5120,0.016520889268981088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,6144,0.020968889196713764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,6144,0.012253333297040729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,8192,0.014564444621404013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,4096,0.014459555347760519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,5120,0.011214222345087262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,5120,0.013793778088357715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,7168,0.01699022286468082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,5120,0.018931556079122756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,3584,0.012740444805887012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,4096,0.009883555273214975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,4096,0.017275555266274344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,3584,0.008998221821255153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,3584,0.011622221933470832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,3584,0.016194666425387066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,3072,0.010701333483060202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,3072,0.008369777765538957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,3072,0.015116444892353482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,7168,0.02288444505797492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,4096,0.012474666866991254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,2560,0.013775111072593264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,2560,0.00796977761718962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,3072,0.011612444288200803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,16384,0.022202667262819078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,2048,0.007471111085679796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,1536,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,2048,0.012787555654843649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,2560,0.010088889135254754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,1024,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,1536,0.007027555671003129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,1536,0.010440000229411654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,10240,0.015936000479592215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,1024,0.00794933322403166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,768,0.004563555535342959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,1024,0.006647999915811751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,1024,0.009547555612193214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,512,0.0038942222793896994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,768,0.0064186664919058485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,768,0.006704000135262807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,768,0.008782222039169734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,256,0.003551111039188173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,512,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,512,0.008373333348168267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,1536,0.009346666435400644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,128,0.00318755561278926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,256,0.005967999911970562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,256,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,2048,0.009176889227496253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,128,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,128,0.008027555214034187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,32,0.0031973332580592898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,2048,0.008390222158696916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,64,0.005630222045713001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,512,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,65536,0.12757155630323622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,32,0.006268444574541516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,256,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,65536,0.11173066827985977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,51200,0.10023378001319037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,65536,0.06556533442603217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,64,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,16384,0.04016710983382331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,51200,0.08642133076985677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,51200,0.05319199959437052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,65536,0.1355466710196601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,12288,0.03359377715322707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,16384,0.03204177816708883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,16384,0.04142933421664768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,12288,0.024663999676704407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,2560,0.009740444521109263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,10240,0.029255112012227375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,12288,0.033320890532599554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,128,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,8192,0.022816888160175745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,10240,0.021406221720907424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,10240,0.02897600001758999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,8192,0.01901511185699039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,8192,0.013779555757840475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,7168,0.020198222663667467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,8192,0.02549955579969618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,7168,0.012827555338541666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,7168,0.01664888858795166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,6144,0.01757066614098019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,7168,0.02258488867017958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,5120,0.014868444866604276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,10240,0.01574844453069899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,6144,0.015392889579137167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,6144,0.020820443828900654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,5120,0.013553777502642738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,51200,0.1079075535138448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,5120,0.018728888697094388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,4096,0.009631111390060848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,3584,0.011725333001878528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,12288,0.017535999417304993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,4096,0.016866667403115165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,3584,0.009016888837019602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,6144,0.011971555650234222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,3584,0.015663110547595553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,4096,0.014163555370436775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,3072,0.010598222414652506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,2560,0.009381333159075843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,3072,0.014647111296653748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,4096,0.012078222301271228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,2560,0.0094035557574696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,2048,0.008760888543393876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,3072,0.010703999963071613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,2048,0.008040000167157915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,16384,0.020986666282018025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,2048,0.012075555821259817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,1536,0.008692444198661381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,2560,0.01367200000418557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,3584,0.011354666617181567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,1024,0.005249777601824866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,1536,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,1536,0.01073066641887029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,5120,0.01053066634469562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,768,0.004552888787455029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,1024,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,1024,0.007363555332024892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,1536,0.007428444094128079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,1024,0.009699555734793345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,512,0.004234666625658671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,768,0.006301333506902059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,768,0.006919999917348226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,256,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,2560,0.007710222568776872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,512,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,512,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,128,0.0032186667538351486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,256,0.006011555592219035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,256,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,128,0.00526577772365676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,64,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,128,0.007666666474607255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,32,0.0029004443850782183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,768,0.009025777379671732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,64,0.005630222045713001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,32,0.00600622221827507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,65536,0.11646400557623969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,256,0.005575110928879843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,65536,0.10752533541785346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,65536,0.06338489055633545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,128,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,3072,0.00817599975400501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,65536,0.13570666313171387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,16384,0.03420533405409919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,51200,0.051071110698911876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,512,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,51200,0.10780000024371678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,16384,0.02048711147573259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,12288,0.02942755487230089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,16384,0.04131733377774557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,12288,0.023756444454193115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,12288,0.016905777984195285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,51200,0.09567733605702718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,2048,0.007341333561473423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,10240,0.02083555526203579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,8192,0.022557333111763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,10240,0.028640889459186133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,16384,0.030598223209381104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,8192,0.018147556318177115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,8192,0.013456000222100152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,8192,0.024762666887707178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,12288,0.03275022241804335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,7168,0.016159999701711867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,7168,0.0125937776433097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,7168,0.022453332940737408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,51200,0.08353600237104629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,10240,0.01512888901763492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,5120,0.01293688929743237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,6144,0.011337777806652917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,6144,0.014815110299322339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,6144,0.020448888341585796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,4096,0.011397333608733283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,7168,0.019613333874278598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,5120,0.018283555905024212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,10240,0.025069332785076563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,4096,0.01200977795653873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,4096,0.016806221670574613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,3584,0.008391111261314815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,3072,0.010395555860466426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,3584,0.01585422290696038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,5120,0.010039110978444418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,5120,0.01330933305952284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,3072,0.009864889085292816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,2560,0.008941333327028487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,3072,0.008014221986134848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,3072,0.014754666222466363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,2048,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,3584,0.010275555153687796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,2560,0.007633777956167857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,2560,0.008355555435021719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,1536,0.006640888750553131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,2560,0.01331911070479287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,2048,0.008007110820876228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,3584,0.011056000159846412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,2048,0.012065778175989786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,1536,0.00700800038046307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,1536,0.01088177743885252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,1024,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,768,0.004564444637960858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,1024,0.009405333134863112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,4096,0.0090524446633127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,6144,0.014666666587193808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,768,0.0063182223174307085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,2048,0.0075031113293435835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,768,0.009030222064918941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,1024,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,256,0.0035297779573334586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,512,0.005961777849329843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,512,0.008006222546100616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,1024,0.006325333482689328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,128,0.0032471112079090546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,256,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,256,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,768,0.006302222195598815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,512,0.004224888980388641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,128,0.005383111122581694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,32,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,128,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,512,0.006249777972698212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,65536,0.0882622202237447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,32,0.005631999837027655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,256,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,65536,0.06542311112085979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,65536,0.06145866711934408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,65536,0.13518933455149332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,64,0.0031999999450312722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,16384,0.029157333903842505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,51200,0.049031111929151744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,51200,0.10754311084747314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,16384,0.02010311186313629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,16384,0.01976088848378923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,16384,0.04043466515011258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,51200,0.07260088788138495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,12288,0.015999999311235216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,128,0.005961777849329843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,10240,0.02163911031352149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,12288,0.016329778565300834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,51200,0.05181599987877739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,8192,0.01587466730011834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,10240,0.01462222304609087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,10240,0.01478844384352366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,64,0.005622222191757626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,10240,0.02868710954984029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,8192,0.013052444491121503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,12288,0.02497333288192749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,8192,0.024823110964563157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,6144,0.0129120002190272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,7168,0.011828444070286222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,7168,0.02199466692076789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,12288,0.03217511044608222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,6144,0.020399111840460036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,6144,0.010712888505723743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,5120,0.01142133358452055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,1536,0.008025777836640676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,7168,0.01424533294306861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,4096,0.01019200020366245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,5120,0.010041777458455827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,5120,0.018393778138690524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,7168,0.011738667057620155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,3584,0.009294222626421187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,4096,0.016238222519556682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,6144,0.011063999599880643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,3584,0.008006222546100616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,3584,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,3584,0.01519555515713162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,8192,0.013463999662134381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,5120,0.009441778063774109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,2560,0.007365333537260692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,3072,0.007841777470376756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,3072,0.013821333646774292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,4096,0.008998221821255153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,2048,0.006602666858169768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,2560,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,2560,0.0069982219073507525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,2560,0.01277511070171992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,1536,0.005214222189452913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,2048,0.007107555866241455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,3072,0.00867022242810991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,2048,0.011443555355072021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,1024,0.004451555510361989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,1536,0.006737777756320105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,1536,0.006646222124497096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,3072,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,1024,0.006610666712125142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,1024,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,1024,0.009418666362762451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,512,0.0035217776894569397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,768,0.009002666506502364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,2048,0.006411555740568373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,512,0.005964444329341252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,256,0.0032746667663256326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,4096,0.008895999855465358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,256,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,256,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,128,0.0028488888508743713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,256,0.008004444340864817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,512,0.00869422240389718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,128,0.005081777771313985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,1536,0.01036800030204985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,64,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,32,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,768,0.0058142224119769196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,32,0.005583999885453119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,512,0.005565333283609814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,65536,0.06407555606630114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,768,0.0042364444169733255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,128,0.008003555238246918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,65536,0.06018400192260742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,65536,0.13561067316267225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,768,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,51200,0.04967911044756571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,16384,0.02667733364635044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,51200,0.04795911245875888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,51200,0.10783733261956109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,12288,0.02095111045572493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,16384,0.019518221418062847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,16384,0.019398222366968792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,16384,0.04031733340687222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,10240,0.018572444717089336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,12288,0.01610133383009169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,128,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,12288,0.03223377797338698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,8192,0.01824977828396691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,10240,0.014431110686726041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,10240,0.013795555465751223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,10240,0.028572443458769057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,65536,0.07878400219811334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,7168,0.015823110938072205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,8192,0.012784889174832238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,8192,0.0247688889503479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,6144,0.014469333820872836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,7168,0.011405333048767514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,7168,0.022095999783939783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,51200,0.06353511412938435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,6144,0.010977778169843884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,6144,0.01074577785200543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,6144,0.020106666617923312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,64,0.005296000176005893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,5120,0.009137777818573846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,5120,0.010020444790522257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,5120,0.018238221605618794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,4096,0.008656000097592672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,3584,0.009031111167536842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,12288,0.015507555670208402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,5120,0.010944888823562197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,3584,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,4096,0.010228444304731157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,8192,0.012522666818565793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,3072,0.007371555599901412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,4096,0.01628266606065962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,3072,0.00776800016562144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,2560,0.007490666376219855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,3584,0.01499288943078783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,3072,0.008094222181373173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,2560,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,2560,0.007595555649863348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,2560,0.01254755589697096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,3072,0.01386044422785441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,2048,0.006292444550328785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,1536,0.005244444641802046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,2048,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,3584,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,2048,0.011437333292431302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,1024,0.004280888785918553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,1536,0.0069439998931354946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,1536,0.006618666566080517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,1536,0.010382222632567087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,1024,0.006311999840868845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,1024,0.009709333380063375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,768,0.0057608890864584185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,768,0.009001777403884465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,768,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,512,0.003519111209445529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,512,0.005286222116814719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,768,0.0038977778620190094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,512,0.005994666781690385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,7168,0.011388444238238864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,256,0.005247999810510212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,128,0.002918222298224767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,256,0.008047111332416534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,128,0.004950222041871813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,128,0.008037333687146505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,128,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,2048,0.006266666783226862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,256,0.003321777822242843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,4096,0.00869422240389718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,64,0.005335111171007156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,256,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,32,0.005312888986534542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,65536,0.05875911315282186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,51200,0.056624001926845975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,64,0.0028977776981062363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,512,0.008857777549160851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,51200,0.04827733172310723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,16384,0.022481777601771887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,65536,0.06951466533872816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,51200,0.10733422305848862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,16384,0.018911111685964797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,16384,0.019333332777023315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,12288,0.018924444913864136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,65536,0.1351591083738539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,1024,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,12288,0.015200000670221118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,10240,0.01642666591538323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,12288,0.03199733297030131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,51200,0.04690666662322151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,10240,0.014144000079896716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,10240,0.013575110998418597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,10240,0.028079112370808918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,8192,0.012081777883900536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,7168,0.013586666848924426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,8192,0.024487111303541396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,12288,0.01590133375591702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,7168,0.011056000159846412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,6144,0.012181333369678922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,7168,0.022171555293930903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,7168,0.010993777877754636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,8192,0.014471999473041959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,6144,0.009935110807418823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,6144,0.010383110907342698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,6144,0.019966221517986722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,65536,0.060722668965657554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,16384,0.0402453343073527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,5120,0.009359999663299983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,5120,0.018216000662909616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,32,0.0028853333658642242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,3584,0.008996444443861643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,4096,0.016132444143295288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,5120,0.010976889067225985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,3584,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,3072,0.008036444584528605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,8192,0.012464889221721225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,5120,0.009409777820110321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,3072,0.00811644477976693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,2560,0.007295110987292395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,3072,0.013445333474212222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,4096,0.008575110799736446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,4096,0.008834666675991481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,2048,0.006587555425034628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,2560,0.0068604445291890045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,2560,0.012306667036480375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,3584,0.008224888808197444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,1536,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,4096,0.009713778065310584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,3072,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,1536,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,1024,0.004577777865860197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,1536,0.010397333237859938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,2560,0.007361777954631382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,1024,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,1024,0.009374221993817223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,768,0.003945777813593547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,1024,0.006575111299753189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,2048,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,2048,0.011239111423492432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,512,0.003847111016511917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,768,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,768,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,768,0.009018667042255402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,256,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,1536,0.00665511108107037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,512,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,512,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,128,0.0028799999919202593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,256,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,256,0.004906666775544484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,256,0.007793777518802219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,3584,0.014346667461925082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,128,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,32,0.0028773333049482773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,128,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,65536,0.059346669250064425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,64,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,32,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,512,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,51200,0.0488337775071462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,65536,0.05741777684953478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,65536,0.1360586616728041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,64,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,16384,0.020409777760505676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,51200,0.10736711157692803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,16384,0.018152889278199937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,16384,0.018959111637539335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,16384,0.04018755422698127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,12288,0.014918223023414612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,2048,0.006454222318198945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,12288,0.03189599845144484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,10240,0.015712888704405892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,10240,0.013426666458447775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,10240,0.028115553988350764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,12288,0.018200000127156574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,8192,0.01370044466522005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,65536,0.05899022022883097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,8192,0.011432888607184092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,8192,0.012096888489193387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,7168,0.013026666310098437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,12288,0.01577422188387977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,51200,0.04570577873124016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,6144,0.011414222419261932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,7168,0.011024889018800525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,7168,0.010433778166770935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,7168,0.02197155521975623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,6144,0.009742221898502773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,5120,0.010033778018421596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,10240,0.01350400017367469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,6144,0.020288000504175823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,4096,0.008893333375453949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,5120,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,5120,0.008778666456540426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,5120,0.0177208897140291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,8192,0.02446488870514764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,3584,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,4096,0.008375999828179678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,4096,0.015511110424995422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,3584,0.008039111064540016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,3072,0.007560888926188151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,3584,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,3584,0.014477334088749357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,2560,0.006654222392373615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,6144,0.010086221827401055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,3072,0.007191110816266801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,3072,0.01309866706530253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,2048,0.00591822216908137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,2560,0.007609777980380588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,2560,0.012433778080675336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,128,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,4096,0.007867555651399825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,1536,0.004903111192915174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,2048,0.007107555866241455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,2048,0.011454222102959951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,1024,0.003929777691761653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,1536,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,1536,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,3072,0.007711999946170383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,768,0.003871110992299186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,1024,0.006335999816656113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,1024,0.00927111092540953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,2560,0.006924444602595435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,768,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,768,0.006309333360857434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,768,0.009039999710188972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,2048,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,256,0.0032151111712058387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,512,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,512,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,512,0.008369777765538957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,128,0.0028924445311228433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,256,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,256,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,256,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,128,0.0052871112194326185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,128,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,64,0.002887111157178879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,32,0.00291377781993813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,51200,0.0461751123269399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,64,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,32,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,1536,0.010383110907342698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,1024,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,65536,0.05615288681454129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,65536,0.1353786653942532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,128,0.00776088900036282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,512,0.003549333247873518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,16384,0.01830222209294637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,51200,0.04511733187569512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,51200,0.04446577694680956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,51200,0.10717511177062988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,12288,0.014863111906581454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,16384,0.039968000517951116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,12288,0.014719110396173267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,51200,0.039832890033721924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,10240,0.01331999980741077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,16384,0.01773066653145684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,65536,0.05647022194332547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,10240,0.013112000293201871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,8192,0.013487999637921652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,12288,0.015140444040298462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,10240,0.01385244478782018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,7168,0.011926222178671094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,8192,0.01108977778090371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,8192,0.02403644389576382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,65536,0.05025066600905525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,7168,0.01038755559258991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,6144,0.01111288865407308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,7168,0.021301334102948506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,16384,0.01843644513024224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,5120,0.01035822182893753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,6144,0.010041777458455827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,10240,0.028117332193586562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,6144,0.019089778264363606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,4096,0.010785777535703448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,5120,0.00904977818330129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,5120,0.017198221551047433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,4096,0.007927999728255803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,4096,0.00851022203763326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,4096,0.01517688896920946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,7168,0.010694222317801582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,3584,0.0081395556529363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,3584,0.007370666497283512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,6144,0.009581333233250512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,5120,0.00850222259759903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,3072,0.007028444773621029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,2560,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,8192,0.011762667033407422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,3072,0.007637333538797166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,2560,0.0074142225914531285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,2048,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,2560,0.012102222277058495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,3072,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,3584,0.014210666219393412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,1536,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,2048,0.007351111206743453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,12288,0.031629333893458046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,3072,0.013180444637934366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,1024,0.0042364444169733255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,1536,0.01036888857682546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,3584,0.010198222266303169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,2560,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,768,0.00388355553150177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,1024,0.006308444258239534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,1024,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,1024,0.009394666386975182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,512,0.003565333369705412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,768,0.0062977779242727495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,2048,0.011409777734014722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,768,0.008665777742862701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,256,0.0032017777363459268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,512,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,1536,0.006004444426960415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,512,0.008343111309740279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,128,0.0032213332338465583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,256,0.005963555640644497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,256,0.007771555748250749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,128,0.004934222333961063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,64,0.0028782222006056043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,2048,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,768,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,64,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,512,0.005292444593376583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,65536,0.03920000129275852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,256,0.005104889058404499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,65536,0.05411110983954536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,51200,0.030466665824254353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,65536,0.13470489448971218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,128,0.008052444292439355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,32,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,16384,0.018222222725550335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,51200,0.042731556627485484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,51200,0.044580443037880786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,51200,0.1068097750345866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,32,0.005122666557629903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,12288,0.015153777268197803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,16384,0.01718577742576599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,16384,0.017809778451919556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,16384,0.03994311226738824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,10240,0.01370844410525428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,12288,0.013757333159446716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,12288,0.014313777287801107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,128,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,12288,0.03162399927775065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,8192,0.012277333272827996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,10240,0.012413333687517377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,10240,0.027107555005285475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,8192,0.011066666907734342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,7168,0.011748444702890186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,8192,0.010775999890433418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,8192,0.024285333024130926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,6144,0.010992888775136737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,7168,0.010031110710567897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,7168,0.021164443757798936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,65536,0.05537955628501045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,6144,0.009295110901196798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,5120,0.009699555734793345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,10240,0.012815111213260226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,5120,0.00903911143541336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,5120,0.008614222208658854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,1536,0.0068853331936730286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,7168,0.010409778190983666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,3584,0.007696000238259633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,4096,0.007742221984598372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,4096,0.008661333057615492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,4096,0.014821334017647637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,3072,0.0069511110583941145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,3584,0.008003555238246918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,3584,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,3584,0.013920888304710388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,6144,0.009686222506894005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,3072,0.007634667058785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,3072,0.012784000072214337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,4096,0.008700444466537898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,2048,0.005327111317051782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,5120,0.016939555605252583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,2560,0.012435555458068848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,2048,0.006330666856633292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,1536,0.0052879999081293745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,2048,0.01108444482088089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,6144,0.01912088857756721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,3072,0.0070257774657673305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,1024,0.00423200014564726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,1536,0.010071111222108206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,2560,0.006804444309737947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,1024,0.005624889085690181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,768,0.003591999825504091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,2048,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,768,0.005797333187527127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,768,0.00902133352226681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,768,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,512,0.0035342222286595237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,1536,0.006719999843173557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,512,0.005920889063013925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,256,0.0032177778581778207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,512,0.00867733359336853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,1024,0.009356444080670675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,256,0.005879111174080107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,128,0.0028773333049482773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,256,0.008010666403505538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,2560,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,128,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,64,0.002549333290921317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,128,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,2560,0.007359999749395583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,32,0.0028444443725877335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,65536,0.02881866693496704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,1024,0.0062977779242727495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,32,0.004966222163703707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,256,0.005079999979999331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,51200,0.025605332520272996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,65536,0.051810665263069995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,65536,0.05480088790257772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,65536,0.13433421982659235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,16384,0.012795555922720166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,51200,0.044341332382626004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,51200,0.04110755523045858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,128,0.005985777825117111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,51200,0.10633689165115356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,16384,0.016168889072206285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,16384,0.016806221670574613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,64,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,10240,0.011659555964999728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,12288,0.013168888787428537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,12288,0.03120000163714091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,10240,0.012086222569147745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,1536,0.005978666659858491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,8192,0.014134221606784396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,12288,0.01035911093155543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,8192,0.010568000376224518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,7168,0.011791110866599612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,8192,0.011176000038782755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,12288,0.01347644461525811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,7168,0.010024888647927178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,7168,0.020638222495714825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,7168,0.01038577738735411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,10240,0.027040888865788777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,6144,0.009374221993817223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,8192,0.023003555006451074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,5120,0.010431110858917236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,5120,0.008340444829728868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,10240,0.012429333395428128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,5120,0.009011555876996782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,4096,0.008325333396593729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,512,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,6144,0.018574222922325134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,16384,0.03973866502443949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,4096,0.007679111427730984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,4096,0.00833777752187517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,4096,0.01517955627706316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,3584,0.007974222302436829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,6144,0.011078221930397881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,3584,0.014149333039919535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,5120,0.016882666283183627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,3072,0.00701155596309238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,3072,0.007667555577225155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,6144,0.009703111317422655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,3584,0.007294221884674496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,2560,0.00665511108107037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,2560,0.007356444166766272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,2560,0.012432888978057437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,3584,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,3072,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,2048,0.006333333336644703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,2048,0.007009777757856581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,2048,0.011454222102959951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,1024,0.004272888931963178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,1536,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,1536,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,1536,0.010402667025725046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,768,0.003924444317817688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,2048,0.00656088896923595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,768,0.005648889061477449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,1536,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,768,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,512,0.0035386664999855887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,2560,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,512,0.0052915554907586836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,3072,0.013125333521101209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,1024,0.005646222167544895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,1024,0.009365333451165093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,256,0.005135111096832487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,128,0.002874666617976295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,256,0.008033778104517195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,768,0.008482666479216682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,128,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,64,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,128,0.008021333151393467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,32,0.0028657778683635923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,256,0.0032248888164758682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,1024,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,64,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,32,0.0052791109515561005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,256,0.005962666538026597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,65536,0.05094933178689745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,65536,0.05475377705362108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,65536,0.1339893341064453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,128,0.005616888817813661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,768,16384,0.012532444463835822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,51200,0.10681511296166314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,512,0.00832799987660514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,512,0.0057760001056724125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,768,12288,0.011023110813564725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,16384,0.016326222154829238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,16384,0.01590666671593984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,16384,0.039552887280782066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,12288,0.013463999662134381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,12288,0.01295111080010732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,12288,0.03053599927160475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,10240,0.011744889120260874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,51200,0.04075288772583008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,768,8192,0.011069333387745751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,768,65536,0.022015111313925848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,8192,0.010607110957304636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,768,7168,0.01036711119943195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,768,10240,0.009733333355850643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,768,51200,0.024737776981459722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,7168,0.009696889254781935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,768,6144,0.009007111191749573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,10240,0.012108444339699216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,51200,0.043755554490619235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,10240,0.02644799980852339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,768,5120,0.008658666577604082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,6144,0.009003555609120263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,6144,0.018580444984965853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,8192,0.01091644416252772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,5120,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,768,4096,0.007635555333561367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,5120,0.016862221890025668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,4096,0.0083155557513237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,7168,0.010104888843165504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,4096,0.01514844430817498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,7168,0.02068444424205356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,6144,0.009691555466916826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,3584,0.013844444519943662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,3584,0.00797333319981893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,768,3072,0.006303999986913469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,5120,0.008997333546479544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,3072,0.0070088886552386815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,4096,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,768,2560,0.005913777897755305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,2560,0.006738666858938005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,3584,0.007359999749395583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,2560,0.012093333734406365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,768,2048,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,2560,0.007328888608349695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,8192,0.023008000519540574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,768,1536,0.004568888909286923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,2048,0.006334222025341458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,2048,0.011401777466138205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,3072,0.007634667058785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,3072,0.013092444174819522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,768,1024,0.0038746665749284956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,1536,0.00665866666369968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,1536,0.010375111467308467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,768,768,0.0037084445357322693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,1024,0.006334222025341458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,1024,0.005640000104904175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,1024,0.009352888498041365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,768,512,0.0033688888781600525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,768,0.0052844443255000645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,768,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,2048,0.006992888947327931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,768,256,0.0028817777832349143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,512,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,512,0.008660444782839881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,768,128,0.0028977776981062363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,256,0.007695111135641734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,128,0.004940444396601783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,768,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,128,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,768,64,0.0029102222373088202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,768,32,0.0029208889851967492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,512,0.005995555470387141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,512,65536,0.022713777091768052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,64,0.005028444445795483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,1536,0.005968889014588461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,65536,0.05029244555367363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,65536,0.0539173318280114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,512,51200,0.01902311046918233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,65536,0.13393955760531956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,128,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,512,16384,0.012815111213260226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,51200,0.040016889572143555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,51200,0.1066124439239502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,16384,0.01624000072479248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,768,3584,0.0069857777820693115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,16384,0.03860088851716783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,12288,0.012806221842765808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,12288,0.03051466743151347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,12288,0.01346933344999949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,512,10240,0.009735110733244155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,51200,0.03974311219321357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,10240,0.011408000356621213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,512,8192,0.009032888544930352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,16384,0.01551555593808492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,8192,0.01035822182893753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,8192,0.02318222158484989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,8192,0.011043555206722684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,7168,0.009682666924264696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,32,0.004943111290534337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,512,12288,0.011075555450386472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,512,6144,0.007704888780911763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,7168,0.01036977767944336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,7168,0.020583111378881667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,6144,0.009144888983832465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,6144,0.009687999884287516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,512,5120,0.008347555167145198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,6144,0.018932445181740653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,5120,0.009071110851234859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,5120,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,512,4096,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,5120,0.016881777180565726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,256,0.005032888717121548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,10240,0.02653955585426754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,512,3584,0.006979555719428592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,4096,0.008364444805516137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,4096,0.007627555893527136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,512,3072,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,3584,0.008047111332416534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,3584,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,3584,0.013847110999955071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,512,2560,0.005618666609128316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,3072,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,3072,0.012851555314328937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,2560,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,2560,0.012441777520709567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,4096,0.015136000182893542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,2048,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,2048,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,3072,0.0069546666410234236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,512,1536,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,2048,0.011341333389282227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,10240,0.012138666378127204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,2560,0.007356444166766272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,512,1024,0.003943110919660992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,1536,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,1024,0.005659555395444234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,1024,0.009023110899660323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,1024,0.0064560001095136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,768,0.005558222118351195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,1536,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,256,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,512,512,0.003532444437344869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,768,0.006257777826653586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,768,0.008722667064931657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,512,0.0052435555391841466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,512,0.005998222364319696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,512,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,256,0.0052800000541739995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,256,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,512,128,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,128,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,512,256,0.0032008888406885993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,512,64,0.00254844439526399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,128,0.005962666538026597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,512,32,0.0028888889484935333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,128,0.008001777860853408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,256,65536,0.018269333574506972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,64,0.004909333255555895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,256,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,32,0.004947555561860402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,256,51200,0.015626667274369132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,65536,0.04895377821392483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,65536,0.13393510712517634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,51200,0.03854666815863715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,512,768,0.003588444242874781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,51200,0.03767022159364488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,51200,0.10571555296579997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,16384,0.015967999895413715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,256,12288,0.01037599974208408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,512,2048,0.005573333137565189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,65536,0.04747466577423943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,12288,0.012784000072214337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,256,10240,0.00942311104800966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,12288,0.03037066592110528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,512,7168,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,256,16384,0.010000000397364298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,256,8192,0.008288000192907121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,10240,0.02641955514748891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,16384,0.01538133290078905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,1536,0.010400888820489248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,8192,0.010869333313571082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,8192,0.010403555300500657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,8192,0.022724444667498272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,7168,0.009376889301670922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,256,6144,0.00849511143234041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,10240,0.012117333710193634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,6144,0.0086986662613021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,16384,0.038726223839653864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,256,7168,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,6144,0.018744000130229525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,5120,0.00833422193924586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,7168,0.020448888341585796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,5120,0.016590222716331482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,10240,0.011402666568756104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,4096,0.007688889073001013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,6144,0.009704888694816166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,256,3584,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,4096,0.01498044364982181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,7168,0.010308444499969482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,3584,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,256,3072,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,256,4096,0.007342221836249034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,3072,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,3072,0.01310488912794325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,256,5120,0.00792266676823298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,4096,0.008345777789751688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,12288,0.0135057775510682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,2560,0.012099555797047086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,3584,0.007978666987684038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,256,2048,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,2048,0.006311999840868845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,2048,0.011071999867757162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,256,2560,0.006321777900060018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,256,1536,0.004583999928500918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,2560,0.006618666566080517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,1536,0.010379555324713389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,3584,0.013802666631009845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,256,1024,0.0038986665507157645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,1024,0.005588444570700328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,1024,0.00907911111911138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,256,768,0.003580444388919406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,2048,0.007003555695215861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,1536,0.005967999911970562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,768,0.009136888715955947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,2560,0.007330666813585494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,768,0.006305777778228124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,512,0.005295111073387994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,512,0.008126222425036961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,3072,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,512,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,256,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,256,0.0076044441925154785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,5120,0.008997333546479544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,256,512,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,256,128,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,256,0.005969777703285217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,128,0.005979555762476391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,128,0.0046471113132105935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,256,64,0.0026355555488003623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,128,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,1024,0.006329777754015393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,128,65536,0.015371556083361307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,64,0.004908444566859139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,32,0.004909333255555895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,65536,0.046219554212358266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,1536,0.006640000061856375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,128,51200,0.0148062225845125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,65536,0.13148444228702122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,51200,0.03767733441458808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,128,16384,0.009486222432719337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,256,32,0.0026151111556424033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,16384,0.01551644504070282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,768,0.005503111001518037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,16384,0.038704888688193426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,16384,0.015832889411184523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,65536,0.04622133241759407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,128,10240,0.008996444443861643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,12288,0.013479111095269522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,12288,0.01256444470749961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,12288,0.030385779009924993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,51200,0.037696000602510236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,10240,0.012088889049159156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,51200,0.10463822550243801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,10240,0.026590221458011206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,128,7168,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,8192,0.011031111081441244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,8192,0.010046222143703038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,8192,0.02271555529700385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,128,12288,0.008332444561852349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,7168,0.009672889278994666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,7168,0.020291555258962844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,6144,0.009653333160612319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,128,5120,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,128,8192,0.008061333662933773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,10240,0.011456888582971362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,256,256,0.003120888852410846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,128,4096,0.006654222392373615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,5120,0.009054222040706212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,5120,0.00831733312871721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,7168,0.010046222143703038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,4096,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,128,3584,0.0063786668082078295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,4096,0.014825777875052558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,6144,0.00903200027015474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,3584,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,128,3072,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,6144,0.01891377733813392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,3584,0.013816888961527081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,128,2560,0.006310222049554189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,3072,0.007650666766696506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,3072,0.01276622215906779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,2560,0.006646222124497096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,2560,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,128,2048,0.006133333262470033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,3584,0.008034666379292807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,2048,0.006343110981914733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,128,1536,0.00452533322903845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,2048,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,3072,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,1536,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,1536,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,1536,0.010395555860466426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,128,6144,0.0073520003093613525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,2560,0.012102222277058495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,128,768,0.003543111185232798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,1024,0.006238222122192383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,1024,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,1024,0.009707555174827576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,768,0.006311999840868845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,768,0.005427555491526921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,2048,0.011071110765139261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,512,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,128,256,0.0028479999552170434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,512,0.00833155545923445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,128,1024,0.0038817777401871155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,5120,0.016915554801623028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,128,128,0.002883555574549569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,256,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,256,0.007709333466158972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,128,512,0.0032400000426504347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,128,64,0.0025128889828920364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,768,0.008386666576067606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,128,32,0.0028524444335036804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,512,0.005250666704442766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,64,65536,0.014879110786649915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,32,0.004587555511130227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,4096,0.007687999970383114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,64,51200,0.013797333670987023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,64,16384,0.00815288888083564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,65536,0.04623466730117798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,64,12288,0.0076657773719893555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,51200,0.03738222188419766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,128,0.004971555537647671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,64,10240,0.0075235557225015424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,12288,0.012788444757461548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,64,8192,0.006794666664467917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,64,0.004934222333961063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,10240,0.011468444433477191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,256,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,8192,0.010295111272070143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,7168,0.00963466697269016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,64,5120,0.006315555423498154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,128,0.005920889063013925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,6144,0.009033777647548253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,64,4096,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,64,3584,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,4096,0.007700444095664554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,64,7168,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,3584,0.007294221884674496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,64,2560,0.006510222123728857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,64,6144,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,64,2048,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,2560,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,64,1536,0.004575110971927643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,5120,0.008336000144481659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,2048,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,64,1024,0.0038862224254343244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,64,768,0.0035866668654812705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,64,3072,0.005997333261701796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,64,512,0.0032417778339650896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,1024,0.005584888988071018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,768,0.005631999837027655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,64,128,0.0028684443483750024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,512,0.0053431110249625305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,64,64,0.002587555597225825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,256,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,64,32,0.002817777709828483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,128,0.004968889057636261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,32,65536,0.014496889379289416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,64,0.004968889057636261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,32,51200,0.013830222189426422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,65536,0.0462719996770223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,64,256,0.0028702221396896574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,32,16384,0.008051555189821456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,51200,0.037321776151657104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,3072,0.006979555719428592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,32,10240,0.007289778027269576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,16384,0.01517688896920946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,16384,0.015140444040298462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,10240,0.011396444506115384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,32,7168,0.006788444601827198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,1536,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,32,6144,0.006451555424266391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,7168,0.009577777650621202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,32,12288,0.008026666939258575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,32,8192,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,5120,0.00835111074977451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,32,0.004579555657174853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,4096,0.007695111135641734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,32,5120,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,32,3072,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,12288,0.012784000072214337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,32,4096,0.0059004442559348206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,3072,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,32,2048,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,8192,0.010033778018421596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,32,3584,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,2048,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,32,1024,0.0035786665976047516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,1024,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,32,768,0.003583111282851961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,2560,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,32,512,0.0031991110493739447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,6144,0.008994667066468133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,32,256,0.002895111011134254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,32,2560,0.006494222415818109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,256,0.004984000076850255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,32,128,0.0028862222615215513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,32,64,0.0025395556456512874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,32,1536,0.004415110995372136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,80,32,32,0.0025395556456512874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,64,0.004939555707905027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,512,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,768,0.005445333404673471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,51200,0.596635553571913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,3584,0.007341333561473423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,32,0.004591111093759537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,16384,0.3862977822621663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,16384,0.19856177435980904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,12288,0.2915777895185682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,16384,0.1333386633131239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,128,0.00488800017370118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,10240,0.21990844938490126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,12288,0.1473368936114841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,51200,0.38988622029622394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,12288,0.09989066918690999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,10240,0.13165688514709473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,10240,0.08437955379486084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,7168,0.15197511514027914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,8192,0.06890133354398939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,6144,0.12989511754777697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,7168,0.06098310814963447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,1536,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,8192,0.1744835509194268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,6144,0.0535662207338545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,5120,0.11255466938018799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,8192,0.10630310906304254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,4096,0.09237066904703777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,5120,0.06627111302481757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,4096,0.03922399878501892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,4096,0.05394222338994344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,3584,0.0786328911781311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,6144,0.08209511306550768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,3072,0.06893689102596708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,3584,0.04843733376926846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,3072,0.031125333574083116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,5120,0.04494400156868828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,51200,1.2113911310831706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,2560,0.03646488984425863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,2048,0.047523554828431874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,3584,0.03470666540993585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,2048,0.03153333399030898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,1536,0.03666844301753574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,7168,0.09102755784988403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,1536,0.018901333212852478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,3072,0.04183377822240194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,1536,0.044179555442598134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,768,0.020639111598332722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,1024,0.014907555447684394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,1024,0.032009777095582746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,512,0.01442399952146742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,768,0.013385777672131857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,768,0.02439466615517934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,256,0.009390222529570261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,512,0.012420444852775998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,512,0.01703733371363746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,256,0.011063111325105032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,256,0.012429333395428128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,2560,0.05743110842174954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,64,0.005245333330498801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,2560,0.03160088923242357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,32,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,64,0.011407111254003314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,128,0.007291555404663086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,32,0.015517334143320719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,1024,0.027708444330427382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,128,0.011050666371981302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,128,0.011402666568756104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,65536,0.62081421746148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,65536,1.1004070705837674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,2048,0.022706665926509436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,51200,0.4701217545403375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,12288,0.20864444308810762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,16384,0.10681955681906806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,16384,0.1580266687605116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,51200,0.32477156321207684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,12288,0.11748977502187093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,10240,0.1748951143688626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,51200,0.8497351010640463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,12288,0.08229866955015394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,8192,0.14256710476345485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,16384,0.27532000011867946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,10240,0.1059911118613349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,7168,0.1259031163321601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,8192,0.05744444661670261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,65536,0.4063413408067491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,6144,0.10807110865910848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,7168,0.05073777834574381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,8192,0.08453955252965291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,6144,0.0632737808757358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,6144,0.04383022255367703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,4096,0.07422666417227851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,5120,0.05351822243796455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,3584,0.06573955880271064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,4096,0.044012443886862866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,3584,0.028992888000276353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,3584,0.03923377725813124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,3072,0.05748889181349012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,7168,0.07156355513466729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,2560,0.04864799976348877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,3072,0.034316443734698825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,5120,0.037752889924579196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,2048,0.039464887645509504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,2560,0.025684444440735713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,4096,0.03271022107866075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,1536,0.03152889013290405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,2048,0.019090667366981506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,2048,0.026141333911154006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,1024,0.023696889479955036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,1536,0.016062221593326993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,1536,0.035565333233939275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,768,0.01700533264213138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,1024,0.013101333545313941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,3072,0.025506666964954797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,1024,0.02573866645495097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,768,0.012182222472296821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,768,0.019165333774354722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,10240,0.06633155875735812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,512,0.014694222145610385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,2560,0.029934220843844946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,128,0.005419555637571547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,256,0.009712888962692684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,256,0.010431999961535135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,128,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,128,0.009758222434255812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,64,0.004683555414279302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,128,0.00998222248421775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,32,0.004910222358173794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,512,0.012239110966523489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,64,0.010031999813185798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,32,0.013129777378506131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,256,0.008477333519193862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,65536,0.22253865665859648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,65536,0.1465324428346422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,16384,0.08326133092244466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,5120,0.09162222014533149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,51200,0.11229244867960612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,12288,0.06523378027809991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,16384,0.040863109959496394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,16384,0.05970488654242622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,65536,0.32574666870964897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,512,0.010817777779367236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,10240,0.05377155542373657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,12288,0.03249511122703552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,51200,0.2492062250773112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,10240,0.02827822168668111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,8192,0.04519199993875292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,12288,0.045215109984079994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,51200,0.17061777909596762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,8192,0.02695288923051622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,6144,0.03446844551298354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,7168,0.021703110800849065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,7168,0.02951644526587592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,6144,0.019592000378502738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,6144,0.025965332984924316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,4096,0.024487111303541396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,5120,0.017625777257813346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,5120,0.022689777943823073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,3584,0.02155911094612545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,4096,0.019153777096006606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,3072,0.019132445255915325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,3584,0.01720533271630605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,8192,0.034280889564090304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,3072,0.013403555585278405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,3072,0.015383111106024848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,5120,0.0291955570379893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,2048,0.013432000246312885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,2560,0.014139556222491793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,2560,0.013780444032616086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,1536,0.011159111228254108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,4096,0.015431111057599386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,2048,0.012350222302807702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,1024,0.008769777913888296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,3584,0.014507555299335055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,1536,0.01497066683239407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,10240,0.038880000511805214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,768,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,1024,0.0083137775460879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,1024,0.012132444315486483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,2560,0.016395555602179635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,768,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,256,0.003926222109132343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,768,0.010114666488435533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,512,0.007325333025720384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,128,0.003543111185232798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,512,0.00831200016869439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,256,0.006667555620272954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,64,0.0032213332338465583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,128,0.0069520001610120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,32,0.0032479998966058097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,1536,0.009372444616423713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,64,0.006578666468461354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,7168,0.03915822174814012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,32,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,512,0.005653333332803514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,65536,0.251893334918552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,51200,0.1974551147884793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,65536,0.12746043999989828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,16384,0.06936711072921753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,51200,0.13094755013783774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,256,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,12288,0.05541333225038317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,128,0.006292444550328785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,16384,0.046837333175871104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,12288,0.02984444300333659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,10240,0.025761778155962627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,8192,0.037913779417673744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,65536,0.17005244890848795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,2048,0.011054221954610614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,51200,0.10234222147199844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,8192,0.026865777042176988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,16384,0.03709422217475043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,7168,0.02277066641383701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,6144,0.02921066681543986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,6144,0.017934223016103108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,10240,0.031144890520307753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,5120,0.025418667329682246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,6144,0.021174222230911255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,4096,0.02198133369286855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,5120,0.01611288885275523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,5120,0.017942221628295053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,8192,0.021889777647124395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,7168,0.03338222371207343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,4096,0.015885333220163982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,4096,0.015547555353906421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,3072,0.016551110479566786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,3584,0.013456000222100152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,12288,0.035959998766581215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,3584,0.01421333352724711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,2560,0.013970666461520724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,3072,0.012457778056462606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,2048,0.012276444170210095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,2560,0.012256888879670037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,2560,0.011745777395036487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,1536,0.010111110905806223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,2048,0.009917333722114563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,2048,0.010808000134097205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,1024,0.008430221842394935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,1536,0.00850222259759903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,1536,0.012967999610635968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,10240,0.045588443676630654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,1024,0.010410666465759277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,768,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,3072,0.01314311143424776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,512,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,768,0.0086666668454806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,256,0.00423200014564726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,512,0.0070186663005087115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,512,0.007376888559924231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,256,0.006689777804745569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,256,0.006000888844331105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,768,0.007073777417341868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,64,0.003576000117593341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,128,0.00598933340774642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,32,0.0035413333939181436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,64,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,32,0.007371555599901412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,65536,0.21320711241828072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,128,0.003920888735188378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,65536,0.16268622875213623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,65536,0.12019556098514134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,128,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,51200,0.12560889456007215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,16384,0.0361422234111362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,12288,0.04596977763705783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,3584,0.018575110369258456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,12288,0.034224000242021345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,7168,0.02369955513212416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,51200,0.0999786655108134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,10240,0.024320888850424025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,8192,0.033960888783137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,51200,0.1699191066953871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,8192,0.026018665896521673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,1024,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,7168,0.029975109630160864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,16384,0.06376444631152682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,6144,0.0287715560860104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,7168,0.019132445255915325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,7168,0.02230666743384467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,5120,0.022696000006463792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,16384,0.04453688859939575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,6144,0.020000888241661918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,4096,0.020636444290479023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,5120,0.01566488875283135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,5120,0.017232000827789307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,3584,0.01754844519827101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,4096,0.014115555418862237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,4096,0.015091554986106025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,3072,0.015052444405025907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,10240,0.03930755456288656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,3584,0.013811555173661975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,2560,0.012993777791659037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,3072,0.012828444441159567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,10240,0.029896001021067303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,2048,0.011403555671374003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,2560,0.011519110865063138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,2560,0.01167377746767468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,1536,0.009374221993817223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,2048,0.009369778136412302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,2048,0.01038577738735411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,1024,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,1536,0.008362666600280339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,1536,0.011695110963450538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,12288,0.027790221903059218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,768,0.006265777680608962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,1024,0.007702222300900354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,512,0.004943999979231092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,1024,0.009109333157539368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,768,0.007277333074145847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,6144,0.01749955614407857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,256,0.0042364444169733255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,512,0.006981333096822103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,128,0.0036026665733920205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,256,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,128,0.006641777853171031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,8192,0.02100444502300686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,128,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,32,0.0033528889632887314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,3072,0.012133333418104382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,64,0.005981333139869902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,65536,0.1751964489618937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,512,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,768,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,256,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,51200,0.133424891365899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,65536,0.0701342225074768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,16384,0.05097511079576281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,64,0.003552888830502828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,51200,0.09137333101696438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,12288,0.037442667616738215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,16384,0.03384888834423489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,32,0.006952889263629913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,10240,0.031007111072540283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,12288,0.026541334059503343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,65536,0.11776710881127252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,10240,0.01643466618325975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,10240,0.023231999741660223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,3584,0.01298577752378252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,7168,0.02310666607485877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,16384,0.023464888334274292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,7168,0.013781333135233985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,6144,0.02028622229894002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,12288,0.01959822244114346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,8192,0.026741334133678015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,5120,0.017720000611411203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,8192,0.01479555500878228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,4096,0.015447111593352424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,5120,0.011603555745548673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,7168,0.017479111750920612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,4096,0.0107360002067354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,4096,0.012620444099108378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,6144,0.0127697777416971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,6144,0.015772443678643968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,51200,0.05704177750481499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,3072,0.010828444527255164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,2560,0.010576888918876648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,3072,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,3584,0.013253333667914072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,2560,0.009031111167536842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,2048,0.009129777550697327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,3584,0.009744889206356471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,2048,0.007663110891977946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,1536,0.007672888537247975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,2048,0.009010666774378883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,1024,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,1536,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,8192,0.020250666472646926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,1536,0.01037599974208408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,768,0.004903111192915174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,1024,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,1024,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,512,0.0042453333735466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,768,0.006607111129495833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,256,0.003526222374704149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,768,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,2560,0.009758222434255812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,512,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,128,0.003306666596068276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,256,0.005719111197524601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,3072,0.012092444631788464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,128,0.005990222096443176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,32,0.0032462223122517266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,5120,0.013963555296262106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,32,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,65536,0.1567280027601454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,512,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,65536,0.11530578136444092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,65536,0.06757511032952203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,3584,0.011739555332395764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,128,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,51200,0.055191109577814736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,51200,0.0899439983897739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,64,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,16384,0.02197244432237413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,16384,0.03263377812173631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,10240,0.029498666524887085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,12288,0.018038221531444125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,12288,0.025464889076020982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,8192,0.025450666745503742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,10240,0.015951111912727356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,10240,0.021879111727078755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,64,0.0029004443850782183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,7168,0.02200444373819563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,16384,0.04662755462858412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,6144,0.01928711103068458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,7168,0.01696711116366916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,12288,0.035516444179746844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,5120,0.01662577854262458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,6144,0.012355555262830524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,6144,0.015635554989178974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,4096,0.01443822185198466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,5120,0.011131555669837527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,5120,0.013801777528391944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,256,0.006050666587220297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,4096,0.010040889183680216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,4096,0.012315555579132505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,3584,0.009199110998047723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,8192,0.01941066649225023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,3584,0.011402666568756104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,2560,0.010302222437328762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,3072,0.00833155545923445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,3072,0.010672888822025724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,2560,0.009703111317422655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,2048,0.008980444735950893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,51200,0.12637955612606472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,2048,0.00832799987660514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,2048,0.007367110914654202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,7168,0.013152889079517789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,1536,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,1536,0.009727111293209923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,2560,0.008104888929261101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,1024,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,3584,0.012741333080662621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,768,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,8192,0.014160888062583076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,1536,0.0076044441925154785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,768,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,512,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,512,0.00657155571712388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,256,0.00563733321097162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,256,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,128,0.0032026666320032547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,128,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,64,0.002874666617976295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,1024,0.008391111261314815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,32,0.0031848889258172777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,768,0.006329777754015393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,32,0.0063582224150498705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,128,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,65536,0.11141244570414226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,65536,0.06506133079528809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,51200,0.10068888796700372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,64,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,16384,0.03866133424970839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,3072,0.011605333122942181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,51200,0.08550844589869182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,65536,0.12522399425506592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,16384,0.021295999487241108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,1024,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,10240,0.028456889920764502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,12288,0.024367999699380662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,10240,0.015498666299713982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,10240,0.021035555336210463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,512,0.0038720000949170855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,51200,0.05278755558861626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,7168,0.020670221911536325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,8192,0.013766222529941134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,16384,0.031543999910354614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,256,0.0035342222286595237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,7168,0.012811555630630918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,12288,0.01738133364253574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,7168,0.016639111770523917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,6144,0.011911110745535957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,6144,0.015136000182893542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,4096,0.013382222089502545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,5120,0.010704889065689512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,8192,0.023062222533755835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,5120,0.013482666677898832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,3584,0.012088889049159156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,4096,0.009367999931176504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,3072,0.010388444695207808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,3584,0.009024000002278222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,8192,0.018581334087583754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,3584,0.011199110911952125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,2560,0.009350222018029954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,6144,0.01776977711253696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,3072,0.01036888857682546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,2560,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,2560,0.009394666386975182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,1536,0.007347555624114142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,2048,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,1536,0.0069751110341813825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,1024,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,1536,0.00906133320596483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,12288,0.033182223637898765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,768,0.004902222090297275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,1024,0.006610666712125142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,3072,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,1024,0.007654222349325816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,512,0.0042239998777707415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,768,0.006330666856633292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,5120,0.01588888963063558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,256,0.003589333345492681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,512,0.005995555470387141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,128,0.0032097777972618737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,256,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,128,0.005995555470387141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,128,0.0053075556125905775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,2048,0.008714666797055138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,64,0.005819555371999741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,768,0.006738666858938005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,2048,0.007333333293596904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,65536,0.11910222636328803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,256,0.005979555762476391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,51200,0.09312266773647732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,65536,0.0627982219060262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,64,0.0028942223224374983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,16384,0.03408266769515143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,51200,0.05071111188994514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,32,0.003206222214632564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,512,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,16384,0.030284444491068523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,32,0.006263999889294307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,12288,0.023520888553725347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,4096,0.012128888732857175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,65536,0.10588533348507351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,51200,0.08307022518581815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,10240,0.0151555554734336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,12288,0.028368890285491943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,8192,0.013449778159459433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,8192,0.018255111243989732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,7168,0.018619555566045973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,12288,0.017072000437312655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,10240,0.025075554847717285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,7168,0.016045333610640634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,6144,0.011376889215575324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,6144,0.014910222755538093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,10240,0.020599999361568026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,16384,0.020294222566816542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,5120,0.010040000081062317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,5120,0.013169777890046438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,4096,0.009010666774378883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,4096,0.011797332929240333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,6144,0.01645511057641771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,3072,0.009333333207501305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,3584,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,3584,0.010814222196737925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,2560,0.011431111229790581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,3072,0.007972444097201029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,3072,0.009858667022652095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,2048,0.008316444853941599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,4096,0.011666666302416058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,2560,0.008753777378135258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,1536,0.0069546666410234236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,2048,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,3584,0.010384889112578498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,7168,0.012494222157531314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,1536,0.0069466663731469055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,1536,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,768,0.005058666484223472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,1024,0.006601777755551868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,1024,0.0070017774899800616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,5120,0.01425333321094513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,512,0.004232888834344016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,768,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,2560,0.007636444436179266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,256,0.0035333335399627686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,512,0.005648889061477449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,2048,0.007297777467303806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,256,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,256,0.005628444254398346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,64,0.0028808888875775864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,128,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,32,0.003212444484233856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,64,0.005616888817813661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,8192,0.021456888980335657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,768,0.006267555471923616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,65536,0.08692000309626262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,51200,0.07201244433720906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,65536,0.06029778056674533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,65536,0.06645600001017253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,16384,0.03129244512981839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,51200,0.04859911070929634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,1024,0.006311999840868845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,128,0.0052968888647026485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,16384,0.01960000064637926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,16384,0.020272889071040683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,10240,0.021472889516088698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,32,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,12288,0.01593866613176134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,8192,0.01882488860024346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,10240,0.014702222413486905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,10240,0.014474666780895658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,128,0.003185777821474605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,8192,0.01309866706530253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,8192,0.01314311143424776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,51200,0.0510062211089664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,6144,0.015575110912322998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,7168,0.011747555600272285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,12288,0.02438666588730282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,6144,0.0107022225856781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,6144,0.011056000159846412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,12288,0.0162764439980189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,4096,0.010618666807810465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,5120,0.010038221875826517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,4096,0.008396444221337637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,4096,0.00885333369175593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,7168,0.016919111212094624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,3584,0.008295999632941352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,3584,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,7168,0.011751111182901593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,2560,0.007637333538797166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,3072,0.007303111255168915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,5120,0.009626666704813639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,2560,0.006928000185224745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,3584,0.009131555755933126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,2048,0.006579555571079254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,1536,0.005029333134492238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,3072,0.007831999825106727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,1536,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,1024,0.004288889053795072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,1536,0.006950221955776215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,2560,0.007650666766696506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,2048,0.006267555471923616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,768,0.0038702223036024305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,1024,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,1024,0.006311999840868845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,512,0.0035342222286595237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,2048,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,768,0.0059262220230367445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,256,0.003195555466744635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,512,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,512,0.005590222362014983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,128,0.0028444443725877335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,256,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,5120,0.013912889692518445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,256,0.005165333135260476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,128,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,32,0.0028444443725877335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,128,0.004970666848950916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,512,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,65536,0.08573333422342937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,768,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,32,0.005653333332803514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,3072,0.00964355551534229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,65536,0.059552000628577344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,51200,0.06874933507707384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,65536,0.06274222003089057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,16384,0.02728355593151516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,12288,0.021568889419237774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,16384,0.019367110398080613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,64,0.0029120000286234748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,12288,0.016194666425387066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,12288,0.015433778365453085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,64,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,10240,0.014469333820872836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,8192,0.01721244388156467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,8192,0.012445333103338877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,7168,0.015896888242827523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,51200,0.04814044303364224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,7168,0.011400889191362592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,16384,0.018894222047593858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,10240,0.01888799998495314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,6144,0.009884444375832876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,6144,0.010669333239396414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,10240,0.01385244478782018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,5120,0.009109333157539368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,5120,0.009700444837411245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,8192,0.01275111072593265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,4096,0.008439111212889353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,3584,0.008951110972298516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,4096,0.00833422193924586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,7168,0.011399110986126794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,3584,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,51200,0.049251556396484375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,3584,0.007625777688291337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,3072,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,5120,0.013209777573744455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,2048,0.005983110931184556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,2560,0.007340444458855524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,2560,0.0069288888739215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,1536,0.004902222090297275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,2048,0.007326222128338284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,2048,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,1024,0.004297777596447202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,1536,0.0069520001610120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,3072,0.00833155545923445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,768,0.0038693332009845306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,6144,0.014967110421922473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,1024,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,2560,0.0074417781498697065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,512,0.0035866668654812705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,3072,0.007257777783605788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,256,0.00319377767542998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,512,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,512,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,256,0.00589155571328269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,128,0.003212444484233856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,256,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,64,0.0028702221396896574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,1536,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,128,0.005738666488064661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,32,0.0028488888508743713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,65536,0.07440000110202365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,1024,0.0063475556671619415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,64,0.0052728888889153796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,32,0.005436444448100195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,51200,0.06043822235531277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,768,0.0063164445261160536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,65536,0.05835377507739597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,65536,0.05989333656099108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,51200,0.04655111167165968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,4096,0.011167999770906238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,12288,0.01924533314175076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,16384,0.018460444278187223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,12288,0.015936889582210116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,10240,0.016585777203241985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,128,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,8192,0.014485332700941296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,10240,0.014071111877759298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,10240,0.013456000222100152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,8192,0.011753777662913004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,7168,0.013767999907334646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,8192,0.012262221839692859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,768,0.00565422202150027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,6144,0.012587555580668978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,7168,0.011056889262464313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,51200,0.047376887665854565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,16384,0.01928622192806668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,5120,0.011038222246699862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,6144,0.009717333647939894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,4096,0.009671111073758867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,6144,0.01036000003417333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,12288,0.015776000089115567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,3584,0.0086986662613021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,4096,0.008378666308191087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,4096,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,3072,0.008310221963458592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,3584,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,16384,0.022636445032225713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,2560,0.007159999675220913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,3072,0.007719111111429002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,7168,0.011032889286677042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,2048,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,2560,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,2560,0.0069368887278768755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,1536,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,5120,0.009362666971153682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,2048,0.006637333167923822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,1536,0.006778666542636023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,3584,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,1536,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,768,0.004087111188305749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,1024,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,512,0.003548444559176763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,768,0.006259555617968242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,3072,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,768,0.00565244464410676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,512,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,512,0.005325333525737126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,128,0.0032151111712058387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,256,0.005889777921968036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,256,0.0053075556125905775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,1024,0.004402666870090696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,64,0.0029048888633648553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,128,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,32,0.0032213332338465583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,128,0.004933333231343163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,65536,0.06450133191214667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,64,0.0052355556852287715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,32,0.005316444569163852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,51200,0.04553066690762838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,51200,0.045695109499825366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,16384,0.020201777418454487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,2048,0.007142222589916653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,16384,0.018941332896550495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,16384,0.017822222577200997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,12288,0.01680355601840549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,12288,0.015554666519165039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,12288,0.014561777313550314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,51200,0.047929777039421924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,1024,0.0063271112740039825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,8192,0.013223999904261695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,10240,0.013456888496875763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,8192,0.011220444407727985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,7168,0.012700444294346703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,10240,0.015159999330838522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,7168,0.011042666931947073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,6144,0.011757333245542316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,65536,0.05795200003517998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,6144,0.010012444522645738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,5120,0.010129777921570672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,8192,0.011981333295504252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,5120,0.008656889200210571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,4096,0.009177777502271865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,7168,0.010416000253624385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,65536,0.05724800295299954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,4096,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,3584,0.008287111090289222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,6144,0.010043555663691627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,3072,0.007928888830873702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,3584,0.008096888661384583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,3584,0.007502222226725684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,3072,0.007706666986147563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,5120,0.009007111191749573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,2560,0.007338666253619724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,2560,0.006945778098371293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,2048,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,256,0.0032097777972618737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,4096,0.007711999946170383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,2048,0.007017778025733099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,2048,0.0063733334342638654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,1024,0.004262222184075249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,1536,0.00630044440428416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,2560,0.006721777634488211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,1024,0.006405333264006509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,3072,0.007116444408893585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,768,0.006172444257471297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,512,0.0035386664999855887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,5120,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,1536,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,512,0.005987555616431766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,512,0.005572444448868434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,1536,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,256,0.006138666636413998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,256,0.005059555586841372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,1024,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,128,0.005974222388532426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,768,0.005631111148330901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,10240,0.013215111361609565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,32,0.0029128889242808023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,256,0.003313777761326896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,65536,0.05290311243798998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,32,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,128,0.003018666679660479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,65536,0.055993775526682533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,51200,0.04155022237035964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,65536,0.05557511250178019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,16384,0.017616889543003507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,51200,0.0451822214656406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,51200,0.04391022192107307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,12288,0.014576888746685453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,16384,0.018569777409235638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,64,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,16384,0.017582222819328308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,10240,0.013371555341614617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,12288,0.014456000592973499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,10240,0.012479999827014076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,768,0.0038977778620190094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,10240,0.01312800000111262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,7168,0.011199110911952125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,8192,0.011122666299343109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,8192,0.01144444445768992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,7168,0.010391111175219217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,7168,0.01074222226937612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,5120,0.010298666854699453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,6144,0.009360888765917884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,5120,0.009023110899660323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,4096,0.009687999884287516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,64,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,6144,0.011844444606039258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,3584,0.009357333183288574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,4096,0.008367111285527548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,6144,0.009974222216341231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,3584,0.00813244448767768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,12288,0.014500444134076437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,3584,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,2560,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,3072,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,3072,0.007057777709431118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,2048,0.006660444455014334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,2560,0.006647999915811751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,128,0.004936000125275718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,1536,0.005757333503829108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,2048,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,2048,0.006331555545330048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,1024,0.005620444400442972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,1536,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,3072,0.008052444292439355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,768,0.003907555507289039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,1024,0.00565422202150027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,8192,0.011745777395036487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,512,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,768,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,768,0.005611555443869696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,256,0.003216000066863166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,512,0.005979555762476391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,512,0.005294222384691238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,128,0.0031759999692440033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,256,0.005056888692908817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,128,0.005684444473849402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,1024,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,5120,0.008660444782839881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,32,0.0028968888024489083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,2560,0.007327111230956183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,65536,0.040692445304658674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,32,0.0052791109515561005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,256,0.005710222240951326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,51200,0.03279377685652839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,65536,0.05525066786342197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,64,0.002867555452717675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,16384,0.014647111296653748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,51200,0.04492533206939697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,51200,0.04197511076927185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,12288,0.015976889265908133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,16384,0.018358222312397428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,4096,0.007669332954618666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,10240,0.01365600029627482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,65536,0.05292000042067634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,12288,0.013952000273598565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,1536,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,128,0.004903111192915174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,8192,0.01107911103301578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,8192,0.01037688884470198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,64,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,16384,0.016881777180565726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,7168,0.01038577738735411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,6144,0.011360000405046674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,7168,0.010040889183680216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,5120,0.010716444088353051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,6144,0.009718221922715506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,6144,0.009177777502271865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,12288,0.014281777871979607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,10240,0.012770666844315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,5120,0.00903733323017756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,4096,0.008341333104504479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,3584,0.00831200016869439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,4096,0.007711999946170383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,3584,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,3072,0.0069422221018208405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,8192,0.012714666624863943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,3072,0.007102222078376346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,4096,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,2560,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,3072,0.00775111135509279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,2048,0.00536088893810908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,2560,0.006783999916579988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,3584,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,2048,0.0064311110311084324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,2048,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,1024,0.004291555533806483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,1536,0.0061048890153567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,1024,0.006310222049554189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,7168,0.012081777883900536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,768,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,1536,0.005305777821275923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,512,0.003551111039188173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,768,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,512,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,256,0.003175111073586676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,2560,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,256,0.004981333182917701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,256,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,128,0.0029057777590221832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,128,0.00565422202150027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,64,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,128,0.004949333353175057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,32,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,1536,0.007033777733643849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,512,0.0052382221652401825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,65536,0.029370665550231934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,32,0.005285333428117964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,51200,0.02416533397303687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,65536,0.0546595553557078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,65536,0.05064622229999966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,16384,0.013073777986897362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,51200,0.04423911041683621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,51200,0.040516446034113564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,12288,0.012479999827014076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,16384,0.016348444753222995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,10240,0.01220266686545478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,10240,0.011801777614487542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,12288,0.01312888910373052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,768,0.003682666768630346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,8192,0.01072266697883606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,10240,0.012111110819710625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,10240,0.011519110865063138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,7168,0.011762667033407422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,8192,0.010831111007266574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,8192,0.010386666489972008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,6144,0.010392888552612728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,7168,0.010047111246320937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,16384,0.016216889023780823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,7168,0.009731555150614845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,5120,0.010410666465759277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,6144,0.009697777529557547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,5120,0.008355555435021719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,64,0.00508977762526936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,4096,0.008650666309727563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,5120,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,4096,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,4096,0.007623111208279927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,3072,0.007072000039948358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,3584,0.007978666987684038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,3584,0.007365333537260692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,2560,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,3072,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,3072,0.006955555743641323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,2048,0.0063786668082078295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,2560,0.007294221884674496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,2560,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,12288,0.013360000318951078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,2048,0.006268444574541516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,6144,0.009000889129108852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,3584,0.007143110864692264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,1536,0.00685422205262714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,1536,0.006213333457708359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,768,0.00423555572827657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,1024,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,1024,0.005629333357016246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,512,0.003591999825504091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,768,0.005569777968857024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,512,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,2048,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,5120,0.009010666774378883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,1024,0.004562666846646203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,128,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,256,0.0057964444988303715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,128,0.0059262220230367445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,64,0.0028799999919202593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,768,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,64,0.004939555707905027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,1536,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,1024,0.005664888769388199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,768,65536,0.023736000061035156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,256,0.004937777916590373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,65536,0.05412355396482679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,65536,0.049639999866485596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,768,51200,0.024990222520298425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,51200,0.043813334570990674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,512,0.00526577772365676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,256,0.0032088889016045462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,16384,0.016021332807011075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,768,12288,0.010710222025712332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,32,0.004937777916590373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,12288,0.013450666434235044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,128,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,768,16384,0.012141333685980903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,32,0.002883555574549569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,768,8192,0.009438222481144799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,10240,0.012104888757069906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,768,7168,0.009183111290136972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,16384,0.01572088897228241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,8192,0.01036888857682546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,7168,0.010236444572607676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,768,6144,0.009000889129108852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,7168,0.009711999860074785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,768,10240,0.010272888673676385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,768,5120,0.008667555948098501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,6144,0.009724444813198512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,768,4096,0.007705777883529663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,5120,0.00905511114332411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,10240,0.011746666497654386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,768,3584,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,4096,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,8192,0.010770666930410596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,768,3072,0.006606222026877933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,3584,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,3072,0.007712889048788283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,768,2560,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,3072,0.0069511110583941145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,6144,0.009005332986513773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,768,2048,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,51200,0.039896888865364924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,2560,0.006798222247097228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,768,1536,0.0046675557063685525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,2048,0.007103111180994246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,2048,0.006302222195598815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,768,1024,0.004007110993067424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,3584,0.007349333001507654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,1536,0.006788444601827198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,1024,0.006637333167923822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,12288,0.012771555946932899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,1024,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,768,512,0.0032151111712058387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,768,0.005997333261701796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,2560,0.007289778027269576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,5120,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,512,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,256,0.005716444303592046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,256,0.004979555391603046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,768,768,0.0036159998012913596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,1536,0.005998222364319696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,128,0.004910222358173794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,128,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,768,32,0.0028808888875775864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,768,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,768,256,0.0029066666546795103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,4096,0.00830844458606508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,32,0.004967111266321606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,65536,0.05337511168585884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,512,51200,0.018944889307022095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,65536,0.0491226679748959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,512,16384,0.01276533305644989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,51200,0.039091553952958845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,51200,0.039376888010236956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,512,12288,0.010024000373151567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,16384,0.015379556351237826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,12288,0.013448889056841532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,64,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,512,10240,0.00905777762333552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,10240,0.01219911128282547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,768,128,0.002624888800912433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,10240,0.011384888655609555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,8192,0.01072266697883606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,512,7168,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,8192,0.01037599974208408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,7168,0.010277333358923594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,512,0.0058337777025169795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,12288,0.012770666844315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,6144,0.009690666364298927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,512,5120,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,5120,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,768,64,0.0029360000044107437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,512,4096,0.007634667058785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,5120,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,512,65536,0.022335999541812476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,512,3584,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,4096,0.008344000412358178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,3584,0.007972444097201029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,512,3072,0.006609777609507243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,16384,0.016211556063758004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,7168,0.009408888717492422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,3072,0.0069919998447100324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,4096,0.007706666986147563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,3072,0.007639110916190677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,6144,0.008999110923873054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,2560,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,512,2048,0.005587555468082428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,2560,0.006680888848172293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,512,1536,0.00490133340160052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,2048,0.00700711127784517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,2048,0.006270222365856171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,512,1024,0.003879111260175705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,1536,0.005988444305128521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,512,6144,0.007635555333561367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,512,768,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,1024,0.006309333360857434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,512,8192,0.00832088871134652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,3584,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,512,0.0052782222628593445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,1536,0.006656888872385025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,512,256,0.003177777760558658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,512,0.005923555543025334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,256,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,1024,0.005619555711746216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,512,128,0.0028560000161329904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,256,0.004944889081848992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,512,64,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,768,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,128,0.005924444645643234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,768,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,256,65536,0.0181031111213896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,128,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,256,51200,0.015659555792808533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,65536,0.046599109967549644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,65536,0.047183109654320605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,256,16384,0.010038221875826517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,51200,0.03678755627738105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,512,2560,0.005626666463083691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,16384,0.015265777707099915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,512,32,0.002883555574549569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,512,512,0.0033528889632887314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,256,10240,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,12288,0.013457777599493662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,12288,0.012764444781674279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,256,8192,0.008512000242869059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,10240,0.012269333004951477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,10240,0.011558222273985544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,51200,0.037351999017927386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,256,7168,0.007679111427730984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,8192,0.010151999692122141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,256,6144,0.008336000144481659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,7168,0.010378667049937779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,256,12288,0.009735999835862054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,6144,0.009683555199040307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,6144,0.009030222064918941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,64,0.004946666873163647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,5120,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,5120,0.008303999900817871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,4096,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,8192,0.01090311093462838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,256,3584,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,4096,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,256,3072,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,16384,0.015858666764365304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,7168,0.009654222263230218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,256,2560,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,3072,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,2560,0.007325333025720384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,256,4096,0.007360888852013483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,256,2048,0.004962666581074397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,2048,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,256,1536,0.004589333302444882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,2048,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,3584,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,1536,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,3584,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,1536,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,256,768,0.003547555456558863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,1024,0.0064026667839950984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,1024,0.005305777821275923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,256,512,0.003216000066863166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,2560,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,768,0.005607999861240387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,256,256,0.0028693332440323303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,512,0.0059262220230367445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,512,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,256,128,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,256,0.00592533333433999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,256,0.005051555732885997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,256,1024,0.0038986665507157645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,128,0.005917333480384614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,256,32,0.002529777793420686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,128,0.0047679998808436925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,128,65536,0.015253333581818474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,64,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,768,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,32,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,128,51200,0.015048000547620984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,65536,0.046329776446024575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,51200,0.03772355450524224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,128,16384,0.009707555174827576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,256,64,0.0025804444319672054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,256,5120,0.008037333687146505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,128,12288,0.009610666996902889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,16384,0.015119999647140503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,32,0.004975999808973736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,128,10240,0.008819555242856344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,65536,0.04518311222394308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,10240,0.0114720000161065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,128,8192,0.008062221937709385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,10240,0.012120000190205045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,3072,0.007686222592989604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,128,7168,0.00867377801073922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,8192,0.010759111079904767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,8192,0.010099555883142684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,7168,0.010038221875826517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,12288,0.013501333693663279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,7168,0.009363555245929295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,128,5120,0.007703999678293864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,6144,0.008683555656009251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,51200,0.036660446061028376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,128,4096,0.0070062221752272705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,5120,0.00868888861603207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,128,3584,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,4096,0.008314666648705801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,16384,0.0159599996275372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,4096,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,128,3072,0.006301333506902059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,3584,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,3584,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,128,2560,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,3072,0.007669332954618666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,12288,0.012817777693271637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,2560,0.007297777467303806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,2560,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,128,1536,0.004570666700601578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,2048,0.006344888773229387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,5120,0.008003555238246918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,128,1024,0.0038924444880750445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,1536,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,1536,0.005737777799367905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,128,768,0.003549333247873518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,1024,0.00638755535085996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,1024,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,128,6144,0.00867377801073922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,128,2048,0.0052933332820733385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,6144,0.009707555174827576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,128,256,0.0028657778683635923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,512,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,2048,0.007120889094140794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,128,128,0.0028853333658642242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,256,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,256,0.005197333378924264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,128,64,0.002867555452717675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,128,0.004909333255555895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,128,32,0.002578666640652551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,64,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,64,65536,0.016188444362746347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,3072,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,32,0.0052675555149714155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,512,0.005311111195219888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,65536,0.04522222280502319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,64,16384,0.00941422250535753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,128,0.005618666609128316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,64,12288,0.00832622249921163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,64,10240,0.007681777907742395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,16384,0.015096889601813423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,128,512,0.0031991110493739447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,12288,0.012465777496496836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,64,7168,0.007487110793590546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,8192,0.010042666561073726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,64,6144,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,7168,0.009447111023796929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,64,5120,0.00666933341158761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,6144,0.00867733359336853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,64,4096,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,768,0.005992888990375731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,5120,0.008216888540320927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,768,0.0052791109515561005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,64,51200,0.01514311134815216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,10240,0.0114702218108707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,64,2560,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,3584,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,64,2048,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,3072,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,64,1536,0.004312889029582341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,2560,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,64,1024,0.00360000009338061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,2048,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,64,768,0.003555555724435382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,1536,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,64,512,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,64,3584,0.005971555494599872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,1024,0.005670222143332164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,64,256,0.0030008889734745026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,4096,0.007369777394665613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,64,128,0.0029599999801980127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,64,64,0.002630222174856398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,512,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,128,0.0046311111913787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,51200,0.036665777365366616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,64,0.004946666873163647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,32,65536,0.016065778003798593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,32,0.004912000149488449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,32,51200,0.015159999330838522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,64,3072,0.0058906666106647914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,768,0.0052933332820733385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,32,16384,0.009026666482289633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,32,12288,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,51200,0.036687112516827054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,32,10240,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,16384,0.015110222829712762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,64,32,0.002576888849337896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,32,8192,0.00832622249921163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,12288,0.012465777496496836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,10240,0.011399110986126794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,32,6144,0.0069573331210348345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,8192,0.010059555371602377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,32,5120,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,65536,0.04550666610399882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,32,4096,0.006309333360857434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,5120,0.008096000386608971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,32,3584,0.005961777849329843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,256,0.005076444397370021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,32,7168,0.007635555333561367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,32,2560,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,3584,0.007318221860461765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,7168,0.009369778136412302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,6144,0.00904177791542477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,2560,0.006607111129495833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,2048,0.00630044440428416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,32,1024,0.0039022221333450745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,1536,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,32,768,0.0035413333939181436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,1024,0.0052826665341854095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,32,512,0.0032186667538351486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,64,8192,0.008011555506123437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,32,256,0.0028986665937635633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,32,2048,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,32,128,0.0028799999919202593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,512,0.005003555367390315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,32,1536,0.004289777742491828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,256,0.004970666848950916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,32,32,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,128,0.004591111093759537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,64,0.004967111266321606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,32,3072,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,3072,0.0069822221994400024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,51200,0.49639198515150285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,51200,1.1286142137315538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,4096,0.007679999702506595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,16384,0.3650080098046197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,32,0.004590222405062782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,768,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,12288,0.27250756157769096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,16384,0.16764356030358207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,65,32,64,0.002522666628162066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,16384,0.14446843994988337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,51200,0.41815556420220273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,12288,0.12483999464246963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,10240,0.23016177283393013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,12288,0.11276977592044407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,8192,0.1857208940717909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,10240,0.08737244208653767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,8192,0.09002221955193414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,7168,0.1656471093495687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,8192,0.08070133129755656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,12288,0.21686932775709364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,7168,0.07565422190560235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,7168,0.06369422541724311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,8192,0.14922222826215956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,7168,0.12724710835350886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,16384,0.2931315633985731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,6144,0.11340888341267903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,5120,0.11733778317769368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,10240,0.10570666525099014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,5120,0.055905779202779136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,5120,0.05138399865892199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,6144,0.06640622350904676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,6144,0.05510310994254219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,10240,0.1796142260233561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,4096,0.0421635541650984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,3584,0.08378755384021336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,51200,0.8653910954793295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,3584,0.041454222467210554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,3584,0.039694223139021136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,3072,0.07389866643481784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,3584,0.0702071123652988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,4096,0.04615822103288439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,2560,0.0628933310508728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,3072,0.03640710976388719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,3072,0.061660442087385386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,4096,0.0774408910009596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,2048,0.04979822370741102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,2560,0.031465778748194374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,2560,0.0543742212984297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,4096,0.09640977780024211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,2048,0.027013333307372198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,1536,0.04016977879736159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,2048,0.023039110832744177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,5120,0.09558133284250896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,1536,0.021876444419225056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,1024,0.028905779123306274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,1536,0.01958488921324412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,2560,0.029133333100212946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,1024,0.016373333003785875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,768,0.02255200015174018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,1024,0.015146666102939181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,1024,0.0314666661951277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,2048,0.046253333489100136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,768,0.013170666992664337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,3072,0.031319111585617065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,6144,0.14102310604519314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,1536,0.03951111104753282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,256,0.009406222237481011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,512,0.022671999202834234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,768,0.013818666338920593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,256,0.009069333473841349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,256,0.01868977811601427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,128,0.006732444382376141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,256,0.011174221833546957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,128,0.008360889222886827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,64,0.004579555657174853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,128,0.011160000330872007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,512,0.0125928885406918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,128,0.016210666961140104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,64,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,512,0.015554666519165039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,32,0.01255733354224099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,512,0.010715555813577441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,65536,1.1101022296481662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,65536,0.5234373410542806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,51200,0.39900710847642684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,51200,0.8612266646491157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,768,0.026927111877335444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,16384,0.2876595656077067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,32,0.004881777697139316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,16384,0.10935378074645996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,16384,0.13437778419918484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,12288,0.21099111768934461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,12288,0.10097599691814846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,65536,0.4175768958197699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,12288,0.09190577930874294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,10240,0.17903555764092338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,16384,0.22949332661098906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,51200,0.32944978608025444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,10240,0.0857235524389479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,12288,0.17969600359598795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,8192,0.0577217804061042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,10240,0.14994666311475965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,7168,0.12654933664533827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,8192,0.12259288628896077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,7168,0.061786664856804736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,8192,0.14655289385053846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,7168,0.05064533485306633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,7168,0.10700533125135635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,8192,0.07239466905593872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,6144,0.04961955547332764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,6144,0.09325777822070652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,10240,0.07690488629870944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,5120,0.04614666766590542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,4096,0.0758462217119005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,65536,0.9327270719740125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,5120,0.07955288887023926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,6144,0.05428711242145962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,4096,0.03229066729545593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,5120,0.09341155820422703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,4096,0.06842755609088473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,3584,0.03431288732422723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,6144,0.10798222488827175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,51200,0.7149839931064181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,3584,0.05962044662899441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,5120,0.03846222162246704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,2560,0.04955110947291056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,3072,0.053256889184316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,3584,0.0674168864885966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,2560,0.02658222284581926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,2048,0.040693332751592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,2560,0.024061333802011278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,3584,0.03020355436537001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,2560,0.04668710960282219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,1536,0.03208533260557387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,2048,0.040114667680528425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,3072,0.02607111136118571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,4096,0.03843199875619676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,1536,0.016220443778567843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,1024,0.02300888962215847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,1536,0.03375733229849074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,1024,0.01367377738157908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,3072,0.05849066707823011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,1024,0.027521777484152053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,768,0.011378666592968835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,2048,0.01963555481698778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,512,0.01240711079703437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,768,0.012095111111799875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,1536,0.018163555198245578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,512,0.011063999599880643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,256,0.00830488900343577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,1024,0.013626666532622443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,768,0.01755288905567593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,2048,0.022710222336981032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,256,0.007833778030342525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,128,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,256,0.017047999633683097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,768,0.02349600030316247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,128,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,64,0.004561777744028303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,128,0.014130666851997375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,32,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,512,0.020202666521072388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,64,0.00814933329820633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,32,0.01070577816830741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,65536,0.3513288762834337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,65536,0.17917777432335746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,128,0.00979377743270662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,512,0.009762666291660732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,51200,0.27831021944681805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,65536,0.14634399943881565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,16384,0.08245599932140775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,51200,0.13579289118448892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,256,0.009766221874290043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,51200,0.11908711327446832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,16384,0.04952177736494276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,12288,0.06467288732528687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,16384,0.042152000798119434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,16384,0.05407466822200351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,12288,0.03750222259097629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,65536,0.18386756049262154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,12288,0.04197777642144097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,51200,0.13965955045488146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,10240,0.032487998406092324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,10240,0.028951999213960435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,3072,0.030537777476840552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,10240,0.036714666419559054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,7168,0.03942666782273187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,8192,0.02459200057718489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,8192,0.028120001157124836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,8192,0.03174933460023668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,6144,0.03408622079425388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,7168,0.022150221798155043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,10240,0.053413334820005626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,7168,0.028038223584493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,6144,0.019916445016860962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,6144,0.021990221407678392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,6144,0.025795555777019922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,5120,0.018605333235528734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,4096,0.025161777933438618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,8192,0.045087112320793994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,5120,0.022647110952271357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,3584,0.02163822286658817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,4096,0.01551822159025404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,4096,0.016162667009565566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,4096,0.02014044423898061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,7168,0.0243520008193122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,3584,0.014748444159825643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,12288,0.03325244453218248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,5120,0.029401779174804688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,2560,0.01627377834584978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,3072,0.013210666676362356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,3072,0.016713778177897137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,5120,0.018930666976504855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,2048,0.013776000175211163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,2560,0.012901333471139273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,2560,0.015444444285498725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,2048,0.01093155559566286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,3584,0.014178666803571912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,1536,0.011043555206722684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,2048,0.014146667387750415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,3584,0.01829600003030565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,1024,0.008385777473449707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,1536,0.009691555466916826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,3072,0.01348888874053955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,1536,0.012896888785892062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,768,0.006949333681000604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,2560,0.011790222591824003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,768,0.006899555524190267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,3072,0.019225777851210702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,768,0.009898666706350114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,512,0.005408888889683618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,1536,0.009455111291673448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,512,0.0059262220230367445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,512,0.008671999805503422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,1024,0.008204444415039485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,1024,0.011504000259770287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,256,0.005647999958859549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,128,0.0035333335399627686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,2048,0.011032889286677042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,768,0.007684444387753804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,128,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,128,0.00565244464410676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,64,0.0035342222286595237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,32,0.0035457776652442086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,256,0.0038844446341196695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,64,0.005978666659858491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,65536,0.2541404565175374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,256,0.006678222368160884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,256,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,65536,0.15476266543070474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,51200,0.1951191160413954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,128,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,1024,0.008040889269775814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,51200,0.11811999479929607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,16384,0.0682142244444953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,32,0.006865777903132969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,51200,0.10375822252697414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,65536,0.14724532763163248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,16384,0.0425946679380205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,16384,0.03769333495034112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,16384,0.04496711161401537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,51200,0.11356177594926621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,12288,0.03126755687925551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,12288,0.035641776190863714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,10240,0.028620445066028174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,10240,0.03132889005872939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,10240,0.025847110483381484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,8192,0.036856889724731445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,8192,0.025009777810838487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,8192,0.02756266627046797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,8192,0.02221600049071842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,7168,0.021683555510309007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,12288,0.05321955680847168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,65536,0.13272533151838514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,7168,0.020434666011068556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,7168,0.024479111035664875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,512,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,6144,0.01923822197649214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,7168,0.0321724447939131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,5120,0.016551110479566786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,5120,0.019935111204783123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,5120,0.016519111063745286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,6144,0.028079112370808918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,12288,0.03315466642379761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,4096,0.014458666245142618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,5120,0.02403822210099962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,4096,0.014575110541449653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,3584,0.013384888569513956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,3584,0.013212444053755866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,4096,0.021006221572558086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,3072,0.015988444288571674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,10240,0.04384533233112759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,3072,0.012430222498046027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,6144,0.018183110488785636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,2560,0.01350044459104538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,6144,0.021927111678653296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,2560,0.011788444386588203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,2560,0.011398221883508893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,2048,0.011766222616036734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,2560,0.014616888430383472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,3584,0.01682400041156345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,2048,0.010386666489972008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,2048,0.013456888496875763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,3584,0.01811733345190684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,3072,0.015476443701320224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,3072,0.012746666868527731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,1536,0.012180444267061023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,1024,0.006680000159475539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,1024,0.010468444062603844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,768,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,1024,0.008003555238246918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,768,0.006121777825885349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,768,0.00903733323017756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,512,0.004818666726350784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,512,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,1536,0.009699555734793345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,512,0.0070222218831380205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,256,0.0035662220584021676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,1024,0.008183110919263627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,1536,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,256,0.005607999861240387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,128,0.0035573335157500375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,256,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,2048,0.010225777824719747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,128,0.0053119998839166425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,128,0.006816888848940532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,512,0.008073777788215214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,4096,0.01775911119249132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,32,0.003585777762863371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,768,0.007632888853549957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,65536,0.22100178400675455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,256,0.006678222368160884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,65536,0.14241154988606772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,65536,0.1264844470553928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,64,0.0035662220584021676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,51200,0.17405066225263807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,1536,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,16384,0.05371910995907254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,51200,0.11146044731140137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,51200,0.10085511207580566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,32,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,65536,0.13520799742804632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,16384,0.04081688986884223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,16384,0.03868622250027127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,16384,0.041518221298853554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,10240,0.035197334157096014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,12288,0.03018933203485277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,12288,0.03155733479393853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,128,0.007696888513035244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,12288,0.033667554457982383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,8192,0.030230220821168687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,10240,0.024497777223587036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,10240,0.027168888184759352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,64,0.0059164443777667145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,8192,0.023523555861579046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,8192,0.021406221720907424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,8192,0.025807999902301367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,12288,0.04286311070124308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,6144,0.022631110416518316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,7168,0.02029333346419864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,7168,0.019720888800091214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,7168,0.02327999969323476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,51200,0.10825866460800171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,5120,0.019917334119478863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,6144,0.01741422216097514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,6144,0.01849688920709822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,6144,0.02141777839925554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,5120,0.016102222932709586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,10240,0.02976800004641215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,7168,0.02607022225856781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,3584,0.014478221535682678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,4096,0.014152889450391134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,4096,0.017690666847758822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,3584,0.013100444442696042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,3072,0.013111111190583972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,4096,0.016556445095274184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,3072,0.012072000238630505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,5120,0.019146665930747986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,2560,0.011129777464601727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,3072,0.012104888757069906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,4096,0.014337778091430664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,3072,0.015480000111791821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,2048,0.009544889132181803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,2560,0.01108000013563368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,2560,0.011052444577217102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,2560,0.014412444498803882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,1536,0.00830666638082928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,3584,0.013092444174819522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,2048,0.009912889036867354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,1024,0.006959111326270633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,2048,0.013459555804729462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,1536,0.008360889222886827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,1536,0.011745777395036487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,768,0.005218666460778978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,1024,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,1024,0.010085333552625444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,768,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,768,0.007355555891990662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,512,0.004219555606444676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,2048,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,3584,0.016229333149062265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,256,0.0037164443896876443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,512,0.007000000112586551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,1536,0.008369777765538957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,512,0.008307555483447181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,128,0.0035617777870761026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,256,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,256,0.007991111112965478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,128,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,64,0.003544888976547453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,128,0.008007110820876228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,768,0.008840888738632202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,32,0.003567111161020067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,64,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,512,0.005792888916201062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,32,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,5120,0.015604444675975375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,256,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,65536,0.09831200043360393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,65536,0.07111910978953044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,128,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,51200,0.05712178018358019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,16384,0.04995200037956238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,65536,0.1351111067665948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,65536,0.17935910489824083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,51200,0.10782133208380805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,16384,0.0410168899430169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,1024,0.007654222349325816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,12288,0.021993777818149988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,12288,0.018631110588709515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,10240,0.028507557180192735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,12288,0.03324710991647508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,51200,0.07329599724875556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,8192,0.024480889240900677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,10240,0.019239111079110038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,10240,0.016750221451123554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,10240,0.029151111841201782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,16384,0.029041777054468792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,7168,0.021300445000330608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,8192,0.014928888943460254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,8192,0.025216889050271776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,7168,0.01386755539311303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,7168,0.01480888823668162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,7168,0.02270577847957611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,6144,0.01349866638580958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,5120,0.016307555966907077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,8192,0.016878222425778706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,51200,0.14449421564737955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,6144,0.018888889087571036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,4096,0.013436444103717804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,5120,0.01146577795346578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,6144,0.012768000364303589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,6144,0.021034666233592566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,4096,0.010862222148312463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,3584,0.013109332985348172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,4096,0.01702311138312022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,3584,0.010411555568377176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,5120,0.018903111418088276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,3584,0.009633777870072259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,3072,0.011696889168686338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,4096,0.01071111112833023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,3072,0.009559999737474654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,3072,0.008912000391218398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,5120,0.012151999606026543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,2560,0.008999110923873054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,2560,0.013776000175211163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,2048,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,2048,0.008311111066076491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,2560,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,12288,0.03373600045839945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,2048,0.007685333490371704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,2048,0.01292711082432005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,3584,0.015933333171738517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,1024,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,1536,0.010778666370444827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,1024,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,3072,0.015053333507643806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,1024,0.009712888962692684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,768,0.005069333232111401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,1024,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,1536,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,768,0.006471111128727595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,768,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,768,0.009296000003814697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,1536,0.007294221884674496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,512,0.0052791109515561005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,512,0.00832088871134652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,256,0.004971555537647671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,16384,0.02310933338271247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,256,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,256,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,512,0.003936888857020272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,128,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,128,0.006233777850866318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,128,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,256,0.0032151111712058387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,64,0.005487111293607288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,32,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,65536,0.15721333026885986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,128,0.0033093332830402586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,65536,0.0867004460758633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,65536,0.06781333022647433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,64,0.0032293332947625052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,51200,0.1256835593117608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,32,0.0035573335157500375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,16384,0.04568799999025133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,51200,0.05500088797675239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,16384,0.02676444417900509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,51200,0.10769066545698379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,12288,0.03584266702334086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,16384,0.04075555668936835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,16384,0.022385777698622808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,1536,0.006037333359320958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,10240,0.026847999956872728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,12288,0.01793688866827223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,512,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,12288,0.032920890384250216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,10240,0.015882665912310284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,65536,0.1348088847266303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,10240,0.028703110085593328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,8192,0.01556799974706438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,7168,0.02016444504261017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,8192,0.02518044412136078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,2560,0.008332444561852349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,7168,0.013213333156373767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,7168,0.013808888693650564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,6144,0.017697778013017442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,51200,0.06692622105280559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,7168,0.022857778602176245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,10240,0.017903111047214933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,5120,0.015069334043396844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,6144,0.012332444389661154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,6144,0.01292266696691513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,4096,0.013445333474212222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,5120,0.011325333681371478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,8192,0.014472888575659858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,5120,0.019023999571800232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,4096,0.01001422190003925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,3584,0.011767999993430244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,4096,0.01685066686736213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,12288,0.020303999384244282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,3072,0.011384888655609555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,3584,0.00998133338159985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,8192,0.023777777949968975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,3584,0.01591466698381636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,2560,0.009642666412724389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,3072,0.00832177781396442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,3072,0.0148062225845125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,2560,0.008770667016506195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,4096,0.010386666489972008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,2048,0.008410666551854875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,2048,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,3072,0.009123555488056606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,2048,0.012859555582205454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,5120,0.011403555671374003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,2048,0.007631999750932057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,2560,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,2560,0.013768889009952545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,1024,0.005130666825506422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,1536,0.01071111112833023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,3584,0.009127111070685917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,1024,0.006654222392373615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,768,0.004563555535342959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,1536,0.007120889094140794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,768,0.00564177789621883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,768,0.0063564446237352155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,512,0.0039048890272776284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,1024,0.005621333503060871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,512,0.005296000176005893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,256,0.0035626664757728577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,512,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,512,0.008443555898136562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,256,0.004954666727119022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,128,0.0032337777730491427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,768,0.00906666699382994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,128,0.004976888911591636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,128,0.007674666742483775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,1024,0.009360888765917884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,64,0.0028782222006056043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,32,0.0032622222271230486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,64,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,256,0.006008000009589725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,32,0.005699555493063397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,65536,0.12573599815368652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,256,0.007747555772463481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,51200,0.0986399981710646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,65536,0.07911466889911227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,6144,0.020503999458418954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,128,0.006008000009589725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,51200,0.06041422155168322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,16384,0.03942222065395779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,65536,0.13464889261457655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,51200,0.05317688981691996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,12288,0.0328764451874627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,16384,0.021428444319301184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,16384,0.02437777817249298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,1536,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,10240,0.028176890479193792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,12288,0.018568888306617737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,65536,0.06491377618577746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,10240,0.016374222106403775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,8192,0.02014755540423923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,10240,0.028639998700883653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,51200,0.10765688949161106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,1536,0.0069751110341813825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,7168,0.021645334031846788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,12288,0.01778133379088508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,12288,0.032664888434939914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,7168,0.01310400002532535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,10240,0.015467555986510383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,6144,0.01945600079165565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,7168,0.02232088810867733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,6144,0.011974222130245633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,6144,0.012136889000733694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,6144,0.02032711108525594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,16384,0.04086577892303467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,8192,0.013850666582584381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,5120,0.010984888507260216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,5120,0.0107022225856781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,5120,0.018287110659811232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,7168,0.01276177747382058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,4096,0.010008888940016428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,3584,0.01275199982855055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,4096,0.016561778055297006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,3584,0.008840888738632202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,5120,0.015765332513385348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,3072,0.009520889156394536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,3584,0.01570133368174235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,3072,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,2560,0.008336000144481659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,4096,0.009391110804345872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,2560,0.007621333003044128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,2560,0.007705777883529663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,3584,0.009406222237481011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,2048,0.007634667058785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,8192,0.02440799938307868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,2048,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,1536,0.006583111153708563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,1536,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,8192,0.01481599940194024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,1536,0.010384000009960597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,1024,0.0047457776963710785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,1536,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,3072,0.0088755554623074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,3072,0.014483556151390076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,768,0.004593777573770947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,1024,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,1024,0.00943288869327969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,768,0.005292444593376583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,768,0.006320888797442119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,512,0.003705777641799715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,768,0.008700444466537898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,2560,0.013444444371594323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,512,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,512,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,256,0.0032213332338465583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,2048,0.007605333295133378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,2048,0.012715555727481842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,128,0.0030453333424197305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,256,0.004959110998445087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,256,0.007661333514584436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,128,0.004934222333961063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,64,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,128,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,32,0.0030177777840031516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,64,0.005274666680230035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,4096,0.014159111513031853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,65536,0.11710045072767471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,1024,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,65536,0.07228622171613905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,128,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,51200,0.09327199723985459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,65536,0.13435555828942194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,51200,0.05465066764089796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,512,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,51200,0.05084266596370273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,16384,0.03340088989999559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,51200,0.10716533660888672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,12288,0.029099557134840224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,16384,0.02036888897418976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,65536,0.06279377804862128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,16384,0.04013155566321479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,12288,0.016873778568373788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,12288,0.0321760012043847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,10240,0.015364444918102689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,10240,0.01520088811715444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,8192,0.020505777663654752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,10240,0.028020444843504164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,256,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,7168,0.019189332922299702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,16384,0.02225688927703434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,8192,0.024262222978803847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,7168,0.012336889074908363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,7168,0.012420444852775998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,6144,0.014575110541449653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,7168,0.022202667262819078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,12288,0.01717066599263085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,5120,0.013173333472675748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,6144,0.011396444506115384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,6144,0.011223110887739392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,6144,0.02023555503951179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,4096,0.011615110768212212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,5120,0.010459555519951714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,8192,0.013346666263209449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,5120,0.018200000127156574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,3584,0.010011555420027839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,4096,0.008685333033402761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,4096,0.009645333720578088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,10240,0.024511999554104272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,3072,0.009222221871217092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,3584,0.008332444561852349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,3584,0.008984000318580205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,3584,0.015501333607567681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,2560,0.007770666645632849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,32,0.005608888963858287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,3072,0.014461333552996317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,5120,0.010087110930018956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,2048,0.007204444044166141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,2560,0.007638221813572778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,2560,0.01333333303531011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,2048,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,1536,0.005665777872006099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,2048,0.011756444142924415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,3072,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,3072,0.007738666401969061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,1024,0.004330666528807746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,1536,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,2560,0.007617777420414819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,1024,0.005590222362014983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,1024,0.0064604443808396654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,2048,0.007129777636792924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,768,0.005445333404673471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,768,0.008977777428097194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,768,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,512,0.0035982223020659555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,8192,0.014194667339324951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,512,0.00526577772365676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,512,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,512,0.005985777825117111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,4096,0.016620443926917184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,256,0.004987555659479565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,128,0.003216000066863166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,1536,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,1536,0.01035644445154402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,128,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,128,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,128,0.007672888537247975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,256,0.003375111147761345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,1024,0.009335111412737105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,64,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,65536,0.08726044495900472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,256,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,32,0.005535999933878581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,256,0.008006222546100616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,65536,0.06534577740563287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,51200,0.0704631143146091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,65536,0.06036000119315254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,64,0.0032346666687064697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,51200,0.05096355411741468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,51200,0.04860888918240865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,32,0.0029093333416514923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,16384,0.01980977753798167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,16384,0.01977333260907067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,12288,0.023167999254332647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,12288,0.01585155559910668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,12288,0.03206311000718011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,12288,0.016170667277442086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,16384,0.029000888268152874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,51200,0.10707733366224502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,10240,0.014197332991494073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,8192,0.016063110695944894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,10240,0.027894222074084814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,8192,0.013021333350075616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,8192,0.023860444625218708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,8192,0.012840888566441007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,65536,0.13435999552408853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,768,0.003927999900446998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,7168,0.011783999701340994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,6144,0.01262399968173769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,7168,0.021983999345037673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,10240,0.014519110321998596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,5120,0.011120888921949597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,16384,0.0398595564895206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,6144,0.010833777487277985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,6144,0.010791999598344168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,4096,0.010377777947319878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,5120,0.00942311104800966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,5120,0.009862222605281407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,5120,0.018014222383499146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,10240,0.01867999964290195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,4096,0.008664000365469191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,4096,0.016248888439602323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,7168,0.01181333346499337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,3584,0.008685333033402761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,3072,0.008354666332403818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,3584,0.015040000279744467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,3072,0.007690666450394525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,6144,0.019941333267423842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,3072,0.014231999715169271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,7168,0.014113777213626437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,3584,0.008850666383902231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,2560,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,2560,0.007358222372002072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,2560,0.012823999755912356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,3584,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,2048,0.006292444550328785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,2048,0.0070675553547011475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,2048,0.011728888584507836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,3072,0.007522666619883643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,1536,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,1024,0.003903111235962974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,1536,0.010468444062603844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,4096,0.008370666868156856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,1024,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,768,0.003915555361244413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,1024,0.009695111049546136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,768,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,768,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,768,0.0090515555606948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,1536,0.006649777707126405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,512,0.005274666680230035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,256,0.0031964443624019623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,512,0.005920889063013925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,512,0.008390222158696916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,2048,0.006609777609507243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,128,0.0028568889117903183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,256,0.00508711114525795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,256,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,1536,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,512,0.0033164444482988785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,64,0.0028231110837724474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,32,0.0029155556112527847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,128,0.007975111405054728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,2560,0.0075093333919843035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,65536,0.07706311013963488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,64,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,32,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,51200,0.06255910793940227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,65536,0.05949778027004666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,256,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,65536,0.13394399483998617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,51200,0.04934933450486925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,128,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,51200,0.10686222049925063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,16384,0.018962666392326355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,16384,0.019254222512245178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,12288,0.020617778102556866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,16384,0.03994133406215244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,1024,0.006344888773229387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,12288,0.015385778413878547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,10240,0.018209778600268893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,12288,0.031890667147106595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,51200,0.04759200082884895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,10240,0.013751111096805997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,8192,0.015492444237073263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,10240,0.014496000276671516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,10240,0.028019554085201685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,8192,0.012748444245921241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,8192,0.012654222548007965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,7168,0.013436444103717804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,8192,0.024015110399987962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,12288,0.016125332978036668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,6144,0.012133333418104382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,7168,0.01145600030819575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,7168,0.011416888899273343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,7168,0.021746666895018682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,5120,0.010640000303586325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,6144,0.010120888551076254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,128,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,6144,0.020050666398472257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,4096,0.009725333087974125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,5120,0.00980444418059455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,5120,0.01816177699300978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,4096,0.008340444829728868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,65536,0.06293600135379367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,16384,0.02632355524433984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,3584,0.007267555428875818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,3584,0.008055110772450766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,3072,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,3584,0.014486221803559197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,3072,0.007755555212497711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,2560,0.007295110987292395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,3584,0.008981333010726506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,2560,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,6144,0.010414222048388587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,2560,0.012454222473833295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,2560,0.007546666595670912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,5120,0.009600888523790572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,2048,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,1536,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,2048,0.011060444845093621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,3072,0.0070159998204973005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,4096,0.015991111596425373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,1024,0.004243555582231945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,1536,0.006683555742104848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,1536,0.010098666780524785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,1024,0.005583111196756363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,2048,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,1024,0.006320888797442119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,1024,0.009402666654851701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,768,0.006302222195598815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,768,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,768,0.00871022211180793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,3072,0.013632000320487551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,512,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,512,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,512,0.008388444781303406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,768,0.0038782221575578055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,256,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,256,0.005882666756709416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,128,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,512,0.003565333369705412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,4096,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,64,0.002847111059559716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,128,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,128,0.005099555684460534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,1536,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,256,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,65536,0.06922488742404513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,32,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,65536,0.060468441910213895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,65536,0.058279110325707324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,51200,0.055007110039393105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,256,0.007640889121426477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,65536,0.13391466935475668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,16384,0.02201955517133077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,51200,0.04740533232688904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,32,0.002924444360865487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,128,0.007637333538797166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,51200,0.10696444246504043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,12288,0.018831110662884183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,16384,0.018251554833518136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,64,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,10240,0.015953777564896476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,16384,0.03973066806793213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,12288,0.016009777784347534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,12288,0.014870221416155497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,8192,0.013829333086808523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,12288,0.031876444816589355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,10240,0.014171555638313293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,2048,0.0063368889192740125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,7168,0.013262222210566202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,8192,0.011789333489206104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,8192,0.012134222520722283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,51200,0.04650400082270304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,7168,0.010718222293588849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,7168,0.02164711058139801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,6144,0.010256888965765635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,6144,0.01960444450378418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,5120,0.01035822182893753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,10240,0.027959111664030287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,8192,0.02369511127471924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,5120,0.009366222553782994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,6144,0.012984889249006907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,4096,0.0092284447617001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,16384,0.019270223047998216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,5120,0.017823111679818895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,4096,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,3584,0.008599110775523717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,4096,0.015488000379668342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,10240,0.013443555268976422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,3584,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,3072,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,3584,0.014534221755133735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,7168,0.011215999722480774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,5120,0.009328000247478485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,3072,0.007672000262472365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,3072,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,3072,0.013542222479979197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,4096,0.008392000363932716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,2560,0.007391110890441471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,2560,0.012429333395428128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,3584,0.007287999822033777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,2048,0.0063075555695427795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,1536,0.004908444566859139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,2048,0.007143999967310164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,2560,0.006640000061856375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,1536,0.005985777825117111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,1024,0.004268444246715969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,1536,0.006643555644485686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,6144,0.009671111073758867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,768,0.0038817777401871155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,1024,0.006305777778228124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,1024,0.005668444352017508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,1024,0.009363555245929295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,512,0.0038275557259718576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,768,0.005581333405441708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,2048,0.011756444142924415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,512,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,512,0.0052675555149714155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,256,0.0033208889265855155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,512,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,1536,0.01037688884470198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,128,0.002846222163902389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,2560,0.006643555644485686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,256,0.004934222333961063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,256,0.008016889293988546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,64,0.002867555452717675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,128,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,32,0.0028977776981062363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,128,0.007984889050324758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,768,0.006299555715587404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,768,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,64,0.0052879999081293745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,32,0.0052844443255000645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,51200,0.04770222306251526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,65536,0.13408533732096353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,2048,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,51200,0.04547911220126682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,128,0.006098666538794835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,51200,0.10687377717759873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,65536,0.05893777476416694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,16384,0.018185777796639335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,16384,0.018962666392326355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,12288,0.016917333006858826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,65536,0.05727110968695747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,12288,0.015503111812803479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,10240,0.014831999937693277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,51200,0.04583377639452616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,10240,0.013117333253224691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,10240,0.013643555343151093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,8192,0.013271999855836233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,65536,0.05793066819508871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,256,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,8192,0.011754666765530905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,12288,0.031848000155554876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,7168,0.012448000411192576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,8192,0.011066666907734342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,7168,0.010407999985747868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,7168,0.01072088877360026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,6144,0.011567999919255575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,7168,0.021819555097156104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,10240,0.027798222170935735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,5120,0.011018666956159802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,6144,0.010051555931568146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,6144,0.0196426659822464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,12288,0.014531556102964612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,4096,0.009043555292818282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,5120,0.00902844468752543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,5120,0.017218665944205392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,8192,0.02362755603260464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,3584,0.00811111099190182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,4096,0.008360000120268928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,4096,0.0075733330514695905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,4096,0.015487111277050443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,3072,0.007224000162548489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,3584,0.007334222396214803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,6144,0.010051555931568146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,16384,0.039665778477986656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,2560,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,3072,0.006950221955776215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,5120,0.008663111262851292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,3072,0.01352533366945055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,2560,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,2048,0.005427555491526921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,2560,0.007349333001507654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,2560,0.012153777811262341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,3584,0.008032889001899296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,2048,0.0069751110341813825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,1536,0.004598222259018156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,3584,0.014175110393100314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,3072,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,1536,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,1024,0.004228444563017951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,1536,0.010439111126793755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,16384,0.020270221763186984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,768,0.003653333418899112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,1024,0.005584888988071018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,1024,0.009368889033794403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,2048,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,512,0.003560888684458203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,1536,0.006648889018429651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,512,0.0052728888889153796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,512,0.006107555495368109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,1024,0.0063288890653186375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,256,0.0032071111102898917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,768,0.005321777943107817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,128,0.0028684443483750024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,256,0.0059164443777667145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,128,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,128,0.007639110916190677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,64,0.00295733329322603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,512,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,32,0.0029360000044107437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,64,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,768,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,65536,0.048615998691982694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,2048,0.011415999796655444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,32,0.005231110999981563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,51200,0.038570665650897555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,65536,0.05602222018771701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,65536,0.13414488898383245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,128,0.0059004442559348206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,16384,0.018608889646000333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,51200,0.04502488838301765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,51200,0.1069413357310825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,256,0.004978666702906291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,16384,0.0173662222094006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,256,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,16384,0.03949955436918471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,65536,0.05518044365776909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,12288,0.014122666584120857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,10240,0.013916444447305469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,12288,0.03183199961980184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,51200,0.04377511143684387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,10240,0.012844444149070315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,10240,0.012867555850081973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,10240,0.027791112661361694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,8192,0.011392888923486074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,16384,0.01891644464598762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,7168,0.011735111474990845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,8192,0.023353776997990076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,7168,0.010869333313571082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,7168,0.010022222167915767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,7168,0.021595555875036452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,8192,0.012823999755912356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,6144,0.009353777600659264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,5120,0.009813333551088968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,6144,0.019275556008021038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,12288,0.014849777022997538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,5120,0.008414222134484185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,5120,0.00905511114332411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,4096,0.008920888933870528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,12288,0.014510222607188754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,6144,0.010701333483060202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,3584,0.007635555333561367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,4096,0.008392000363932716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,4096,0.007535999847782983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,4096,0.015219555960761176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,3072,0.006649777707126405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,3584,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,6144,0.009704888694816166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,3584,0.013781333135233985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,2560,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,768,0.008995555341243744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,3072,0.007672888537247975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,3072,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,2048,0.005590222362014983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,5120,0.017550221747822232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,2560,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,2560,0.012079111403889127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,1536,0.004561777744028303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,2048,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,2048,0.011086222198274402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,3584,0.008028444316652086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,1536,0.006945778098371293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,1536,0.010390222072601318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,8192,0.011229332950380115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,1024,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,3072,0.013095999757448832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,2560,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,1024,0.009019555317031013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,2048,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,768,0.009026666482289633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,768,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,1536,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,512,0.0052728888889153796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,512,0.008031110796663497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,256,0.003216888962520493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,512,0.006004444426960415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,768,0.005473777651786804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,256,0.004979555391603046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,128,0.0029377777957253983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,256,0.005767999837795894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,512,0.003539555602603488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,1024,0.00424799985355801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,64,0.0028862222615215513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,128,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,128,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,128,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,1024,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,65536,0.0388355553150177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,32,0.004983110974232356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,65536,0.052925331725014575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,65536,0.05510310994254219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,51200,0.030920889642503526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,65536,0.13355110751258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,16384,0.014929778046078153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,51200,0.041836443874571055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,51200,0.10660977496041192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,768,0.0038853333228164246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,12288,0.013223110801643796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,16384,0.016872889465755887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,16384,0.039319111241234675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,12288,0.014655111564530266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,12288,0.014127110441525778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,10240,0.011981333295504252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,12288,0.03133422136306763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,10240,0.01238044434123569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,10240,0.01237422227859497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,32,0.002846222163902389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,10240,0.027430221438407898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,64,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,8192,0.011057777537239922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,8192,0.01036800030204985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,8192,0.023386667172114056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,7168,0.009861333502663506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,6144,0.009704888694816166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,7168,0.020962667134073045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,51200,0.04455999864472283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,8192,0.0109120003051228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,5120,0.008962666822804345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,6144,0.009715555442704095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,7168,0.010332444475756751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,16384,0.01721600029203627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,4096,0.008681777450773451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,5120,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,5120,0.016858667135238647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,7168,0.010381333529949188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,3584,0.008004444340864817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,4096,0.0074942219588491656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,4096,0.00830844458606508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,4096,0.014502222339312235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,3072,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,3584,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,6144,0.00888355573018392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,3584,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,6144,0.018601778480741713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,3584,0.014135110709402295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,3072,0.007694222033023834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,3072,0.0069324444565508105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,2048,0.0052426668504873914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,3072,0.013103110922707452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,2560,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,2560,0.012103999654452005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,1536,0.004590222405062782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,2048,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,2048,0.006167111297448476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,256,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,2048,0.011050666371981302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,1024,0.004261333495378494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,1536,0.006959999601046245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,1536,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,768,0.0035413333939181436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,1024,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,1024,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,1024,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,512,0.003546666767862108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,768,0.006247111078765657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,2560,0.007294221884674496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,768,0.008664888640244802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,256,0.0032195556494924757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,512,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,512,0.005299555758635203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,512,0.008460444708665213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,128,0.0028702221396896574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,256,0.005712888720962737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,256,0.0074480002125104266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,1536,0.01036800030204985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,64,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,2560,0.005633777628342311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,128,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,768,0.005439110928111606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,64,0.004908444566859139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,5120,0.00906488878859414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,65536,0.030798223283555772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,32,0.0052764444715446895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,65536,0.05052266518274943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,256,0.004953777624501122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,51200,0.025391111771265667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,65536,0.13355199495951334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,128,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,51200,0.044103112485673696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,51200,0.04015200005637275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,32,0.002557333351837264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,16384,0.01640266676743825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,12288,0.010478222535716163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,65536,0.05445955528153313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,12288,0.013110222087966071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,128,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,16384,0.012872888810104795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,12288,0.031249778138266668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,51200,0.10654489199320476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,16384,0.017260443833139207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,16384,0.038951999611324735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,10240,0.027069333526823256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,8192,0.010017777482668558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,7168,0.009290667043791877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,8192,0.022673777408070032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,12288,0.013479111095269522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,10240,0.010641777680979835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,6144,0.0083155557513237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,7168,0.009637333452701569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,7168,0.020364445116784837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,10240,0.012119111087587146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,6144,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,6144,0.018927110566033255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,8192,0.010761777559916178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,5120,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,4096,0.007416889071464539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,5120,0.009177777502271865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,7168,0.010379555324713389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,4096,0.007324444750944774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,3584,0.006953777538405524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,4096,0.01516888870133294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,10240,0.011511111425028907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,5120,0.008371555142932469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,3072,0.006312888943486744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,3584,0.008016889293988546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,3584,0.0070017774899800616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,3584,0.014140443669425117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,2560,0.0059262220230367445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,3072,0.006981333096822103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,5120,0.016879111528396606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,3072,0.012772444221708508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,2048,0.0052391112678580815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,2560,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,2560,0.012429333395428128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,4096,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,2048,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,1536,0.004582222137186262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,2048,0.0070053330726093715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,2048,0.011391110718250275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,1024,0.0038942222793896994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,1536,0.006686222222116258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,1536,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,3072,0.007671111159854465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,768,0.0035573335157500375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,1024,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,1024,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,1024,0.009709333380063375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,512,0.003228444399105178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,768,0.0052782222628593445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,768,0.00869066682126787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,8192,0.009667555491129557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,512,0.0052977779673205475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,256,0.0032417778339650896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,512,0.008055999875068665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,256,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,128,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,256,0.007661333514584436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,2560,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,128,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,64,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,128,0.00777422222826216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,32,0.0028693332440323303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,512,0.005902222047249476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,768,65536,0.025810667210155066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,64,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,32,0.00519822206762102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,1536,0.010382222632567087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,768,51200,0.019945777124828763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,65536,0.054435554477903575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,65536,0.13303644127315944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,51200,0.04007644454638163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,128,0.005969777703285217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,768,16384,0.011402666568756104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,6144,0.008750222623348236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,51200,0.10640622509850396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,768,12288,0.009864889085292816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,16384,0.016057777735922072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,16384,0.03854488995340135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,256,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,65536,0.04928266670968798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,12288,0.012780444489585029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,768,10240,0.010081777969996134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,12288,0.013470222552617392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,768,0.006274666637182236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,12288,0.030798223283555772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,768,8192,0.009012444151772393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,10240,0.012127111355463663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,10240,0.011517333487669626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,768,7168,0.008370666868156856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,8192,0.01074222226937612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,8192,0.010039110978444418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,8192,0.022620444496472675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,16384,0.015523556205961438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,7168,0.010219555762079027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,7168,0.009732444253232744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,7168,0.02067733307679494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,768,5120,0.008360889222886827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,6144,0.009703999592198266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,6144,0.00886222223440806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,6144,0.018575999471876357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,768,4096,0.007646222081449296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,5120,0.008038221961922115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,51200,0.04351111253102621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,10240,0.026791110634803772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,768,3584,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,4096,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,4096,0.015209777487648858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,3584,0.007019555403126611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,3584,0.008029333419269985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,3584,0.014058666096793281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,3072,0.006847111301289664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,3072,0.01310755560795466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,768,2560,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,4096,0.008356444537639618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,768,6144,0.007756444315115611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,2560,0.007324444750944774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,2560,0.012099555797047086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,768,3072,0.007000000112586551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,2048,0.0069875551594628235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,768,1536,0.004581333448489507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,2048,0.011394666300879585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,5120,0.016892444756295946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,3072,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,768,1024,0.0039057777159743835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,1536,0.005607999861240387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,1536,0.006681777950790193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,2560,0.006607111129495833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,768,2048,0.005155555489990446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,768,768,0.003543111185232798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,1024,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,2048,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,768,512,0.0032266666077905228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,5120,0.00904266701804267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,512,0.005227555417352253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,1536,0.010057777994208867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,512,0.00831822223133511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,768,256,0.0031911111954185697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,512,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,1024,0.0064204442832205035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,768,128,0.0026888888743188647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,256,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,256,0.008008889026112026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,768,0.005611555443869696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,768,64,0.0025235555238193935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,128,0.00591111100382275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,768,32,0.0025253333151340485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,128,0.0074542222751511475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,512,65536,0.019075555933846366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,64,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,32,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,65536,0.04876177840762668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,512,51200,0.01575644479857551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,1024,0.005618666609128316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,65536,0.13303555382622614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,768,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,512,16384,0.01069155583779017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,51200,0.03936000002755059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,51200,0.1056213312678867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,16384,0.0151582227812873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,16384,0.01616533266173469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,512,12288,0.009196444518036312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,65536,0.05371288789643181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,12288,0.012517333030700684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,256,0.0052177777720822235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,512,10240,0.009372444616423713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,12288,0.01346666696998808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,768,0.00904977818330129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,128,0.005010666532648934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,512,8192,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,10240,0.02682311170630985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,16384,0.03805333375930786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,8192,0.010155555274751451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,12288,0.030171556605233088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,7168,0.009709333380063375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,51200,0.04084177811940511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,10240,0.012150222228633033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,7168,0.020308444897333782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,6144,0.00868622213602066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,512,7168,0.007704888780911763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,6144,0.009656888743241629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,6144,0.018582221534517076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,7168,0.010381333529949188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,5120,0.008026666939258575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,5120,0.009000889129108852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,512,4096,0.007339555356237624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,5120,0.016958221793174744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,8192,0.010760000182522668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,4096,0.008360000120268928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,512,5120,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,512,3584,0.007019555403126611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,4096,0.015187554889255099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,512,6144,0.007515555454625024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,512,3072,0.006666666517655055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,3584,0.01389955480893453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,3072,0.006811555474996567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,8192,0.022650667362742957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,3072,0.013340444200568728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,3072,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,4096,0.007348444726732042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,2560,0.006649777707126405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,512,2048,0.006303999986913469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,2560,0.012054222325483957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,3584,0.008051555189821456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,512,1536,0.004281777888536453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,2048,0.006991110742092133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,2048,0.006331555545330048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,2048,0.011404444773991903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,512,1024,0.003916444463862313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,1536,0.006029333505365584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,512,2560,0.006490666833188798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,1536,0.010058666268984476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,512,768,0.00360355567600992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,1024,0.006270222365856171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,1024,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,2560,0.007339555356237624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,1024,0.009703999592198266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,512,512,0.0032222221295038858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,768,0.006262222097979651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,768,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,3584,0.007000889215204451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,512,256,0.0032195556494924757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,768,0.009005332986513773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,512,0.005976000179847081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,512,128,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,256,0.005962666538026597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,256,0.004941333499219683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,1536,0.006649777707126405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,512,64,0.0029004443850782183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,128,0.005611555443869696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,512,32,0.002830222249031067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,128,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,256,65536,0.015535111228624979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,64,0.004971555537647671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,10240,0.011412444214026133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,32,0.00463377767139011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,256,51200,0.013816888961527081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,65536,0.04826577835612827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,512,0.005315555466545953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,512,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,256,16384,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,51200,0.03734222385618422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,256,0.007552000383536021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,128,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,51200,0.10408711433410645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,256,12288,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,16384,0.015841777125994366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,16384,0.014860444598727755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,256,10240,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,12288,0.013467555244763693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,12288,0.012652444342772165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,12288,0.030533333619435627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,256,8192,0.0069982219073507525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,65536,0.047008888589011304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,10240,0.026073777013354834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,65536,0.13207822375827366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,256,7168,0.0069822221994400024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,8192,0.010040000081062317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,8192,0.02237333357334137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,7168,0.009648888475365108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,256,6144,0.006653333289755716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,16384,0.03800888856252035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,10240,0.012075555821259817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,10240,0.01128799964984258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,256,5120,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,6144,0.009699555734793345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,8192,0.010830221904648675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,5120,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,5120,0.016533333394262526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,7168,0.020590222544140287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,4096,0.007420444654093848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,6144,0.008967999782827165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,256,3584,0.0063288890653186375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,4096,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,51200,0.038161777787738375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,3584,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,256,3072,0.006313777632183499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,5120,0.008691555923885768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,256,4096,0.006607999818192587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,3072,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,3072,0.013131555583741931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,6144,0.01824977828396691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,256,2560,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,2560,0.0064026667839950984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,2560,0.011755555868148804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,7168,0.010093332992659675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,2560,0.007272000114123027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,2048,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,2048,0.011066666907734342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,256,1536,0.004577777865860197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,1536,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,3584,0.014154665999942355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,256,2048,0.006597333484225803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,1536,0.006663110935025745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,1536,0.010053333308961656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,1024,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,256,768,0.0035546666218174826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,2048,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,768,0.0053102220926019884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,768,0.009018667042255402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,768,0.006252444452709622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,512,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,1024,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,1024,0.008984888593355814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,256,256,0.0028844444702068963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,4096,0.014534221755133735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,512,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,256,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,256,0.00573333352804184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,256,0.00798044436507755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,256,512,0.0033742222521040174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,128,0.005671999934646819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,128,0.004951999833186467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,128,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,256,32,0.002571555475393931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,3072,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,128,65536,0.014509333504570855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,64,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,32,0.00489777781897121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,65536,0.04557777775658501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,3584,0.008027555214034187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,65536,0.04632800155215793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,65536,0.13025511635674372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,51200,0.037695109844207764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,128,16384,0.008034666379292807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,51200,0.10433244705200195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,512,0.0057760001056724125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,16384,0.01483022173245748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,16384,0.015843555331230164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,16384,0.038041777080959745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,256,128,0.0029066666546795103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,128,51200,0.01293244461218516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,128,10240,0.007442666424645319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,12288,0.012434666355450949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,256,64,0.002522666628162066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,51200,0.03668177790111966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,10240,0.011446221835083432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,10240,0.012090666426552666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,10240,0.026259554757012263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,8192,0.010728889041476779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,8192,0.010064889159467485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,8192,0.022387555903858606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,256,1024,0.0038559999730851916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,7168,0.009690666364298927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,128,6144,0.006674666785531574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,12288,0.03030311067899068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,128,8192,0.007078222102589077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,6144,0.008765333228641087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,6144,0.01869155632125007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,128,12288,0.007650666766696506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,5120,0.0081404447555542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,12288,0.01313688854376475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,128,4096,0.005963555640644497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,7168,0.020375111036830477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,5120,0.009012444151772393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,4096,0.008310221963458592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,4096,0.007305777735180325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,4096,0.0151582227812873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,6144,0.009626666704813639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,128,7168,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,128,3072,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,3584,0.006989333364698622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,3584,0.013917333549923368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,3072,0.00666844430896971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,3072,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,3072,0.012780444489585029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,128,3584,0.00630044440428416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,2560,0.011769777370823754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,128,5120,0.0063164445261160536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,128,2048,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,3584,0.008031110796663497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,2048,0.006311111152172089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,2048,0.0070159998204973005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,2048,0.011077333655622272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,128,2560,0.006263999889294307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,1536,0.006640000061856375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,1536,0.005991111199061076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,128,1024,0.003552000141806073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,1536,0.009896888501114314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,2560,0.007292444507280986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,1024,0.005650666852792104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,128,768,0.0035297779573334586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,1024,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,1024,0.009030222064918941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,128,1536,0.004249777644872665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,5120,0.016873778568373788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,768,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,768,0.005427555491526921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,768,0.008664000365469191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,512,0.005981333139869902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,512,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,2560,0.006618666566080517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,256,0.00526577772365676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,256,0.005611555443869696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,128,128,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,256,0.008016000191370646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,128,0.005881777654091517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,128,64,0.002530666689078013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,7168,0.010035555395815108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,128,0.0070088886552386815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,128,256,0.0030737777964936364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,64,0.004933333231343163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,64,65536,0.01071200023094813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,64,51200,0.010049777726332346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,512,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,64,16384,0.007305777735180325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,64,12288,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,65536,0.045203556617101036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,51200,0.036221332020229764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,16384,0.01482222146458096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,64,8192,0.006614222294754452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,12288,0.012440889245933957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,64,7168,0.00664444433318244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,128,512,0.0032071111102898917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,10240,0.011074666347768573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,64,6144,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,64,5120,0.0063235556913746735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,7168,0.009351111120647853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,64,4096,0.005965333431959152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,6144,0.00869599978129069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,64,3584,0.005977777971161737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,5120,0.008159110943476358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,64,3072,0.005990222096443176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,4096,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,64,2560,0.006637333167923822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,3072,0.0070062221752272705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,64,2048,0.005981333139869902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,128,32,0.002573333266708586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,64,1536,0.004580444345871608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,8192,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,64,1024,0.00388355553150177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,2048,0.006351111249791251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,64,768,0.003549333247873518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,1536,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,64,512,0.00319377767542998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,1024,0.005433777968088786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,64,256,0.0028773333049482773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,64,10240,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,768,0.005583999885453119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,3584,0.006989333364698622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,64,64,0.002529777793420686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,128,0.004952888935804367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,256,0.004951999833186467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,128,0.005026666654480828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,2560,0.006574222197135289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,32,65536,0.010752889017264048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,64,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,32,51200,0.010209778116808997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,32,16384,0.007337777978844113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,32,0.0046160001721647055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,32,12288,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,65536,0.04514044523239136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,32,10240,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,16384,0.015039111177126566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,32,8192,0.006659555352396435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,512,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,10240,0.011121778024567498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,32,7168,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,64,32,0.0028275555620590844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,8192,0.010095111197895473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,32,5120,0.0063386667105886675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,7168,0.009388444324334463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,32,4096,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,6144,0.00868622213602066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,32,3584,0.006324444380071428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,5120,0.00815466625822915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,32,3072,0.0063271112740039825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,4096,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,64,128,0.0026142222599850762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,3072,0.00699733363257514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,32,6144,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,32,2048,0.006015111174848344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,32,1536,0.004546666724814309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,32,0.0052133335007561584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,32,1024,0.003954666770166821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,1536,0.005977777971161737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,51200,0.036102222071753606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,32,768,0.003359111025929451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,32,2560,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,32,512,0.0032026666320032547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,768,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,32,256,0.002628444383541743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,2560,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,32,128,0.002752888947725296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,512,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,32,64,0.002628444383541743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,256,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,2048,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,64,32,32,0.0025271111064487034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,128,0.004902222090297275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,64,0.004902222090297275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,32,0.004931555440028508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,12288,0.012432888978057437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,16384,0.3686399989657932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,51200,0.4941706657409668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,16384,0.16631910536024305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,51200,1.1279200447930229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,12288,0.2789955668979221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,16384,0.1350897815492418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,10240,0.22730310757954916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,3584,0.007060444189442529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,51200,0.4266017807854547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,8192,0.18410666783650717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,12288,0.11444621615939671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,10240,0.10503911309772068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,10240,0.09234755569034153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,7168,0.1624222199122111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,8192,0.08867466449737549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,6144,0.13781155480278864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,7168,0.061879999107784696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,5120,0.11978044774797227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,6144,0.05553422371546427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,1024,0.00565155554148886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,4096,0.09472178088294135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,5120,0.05571733580695259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,3584,0.0834311114417182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,4096,0.040122667948404946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,4096,0.04641866683959961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,3072,0.07296978102789985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,3584,0.035700443718168474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,8192,0.0711688862906562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,7168,0.07504355245166354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,3072,0.031515555249320135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,3072,0.03645688957638211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,2048,0.05029155479537117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,2560,0.02748888896571265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,2560,0.032362666394975446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,2048,0.0233608881632487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,2048,0.027612444427278306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,1536,0.018576888574494254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,1536,0.022071111533376906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,1024,0.02888444397184584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,3584,0.04133066534996033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,768,0.022056889202859666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,12288,0.12411555978986953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,1024,0.014907555447684394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,1024,0.01625155574745602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,768,0.01351466692156262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,768,0.013372444444232516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,256,0.009314667019579146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,1536,0.03932888971434699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,256,0.010885333021481832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,256,0.009030222064918941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,2560,0.06280355321036445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,6144,0.06605777475568983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,128,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,64,0.007318221860461765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,32,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,5120,0.046648889780044556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,512,0.01218844453493754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,512,0.010830221904648675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,32,0.012420444852775998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,128,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,128,0.010910222099887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,65536,1.0903600056966145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,51200,0.8493786917792426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,51200,0.3967146608564589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,64,0.009700444837411245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,16384,0.27646756172180176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,12288,0.2078711059358385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,65536,0.40412267049153644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,16384,0.1098888913790385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,65536,0.5216835339864095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,10240,0.17513599660661486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,10240,0.08512355883916219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,12288,0.10011644495858087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,8192,0.1414524449242486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,10240,0.0770346654786004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,7168,0.12534222337934706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,8192,0.0571848882569207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,8192,0.07215822405285306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,6144,0.10636444224251641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,7168,0.050604446066750415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,5120,0.0910151137246026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,6144,0.04471200042300754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,512,0.015495111544926962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,6144,0.05405422382884555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,4096,0.07354044251971774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,51200,0.33291376961602104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,5120,0.04575022061665853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,3584,0.06565066840913561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,4096,0.03796089026663039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,4096,0.03259822063975864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,3072,0.05728089147143894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,3584,0.029651555750105116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,3584,0.03404088815053304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,2560,0.04823733369509379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,3072,0.026447110705905493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,16384,0.13348978095584446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,2048,0.03871377640300327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,2560,0.023052444060643513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,2048,0.019696000549528334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,1536,0.031070222457249958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,7168,0.061386664708455406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,12288,0.0822862254248725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,1536,0.018167111608717177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,5120,0.038457777765062116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,1536,0.016359999775886536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,768,0.01715022159947289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,1024,0.013544888959990608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,2048,0.022925333844290838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,512,0.012454222473833295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,768,0.012079999678664737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,768,0.011706666813956367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,256,0.008354666332403818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,512,0.010761777559916178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,512,0.009974222216341231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,3072,0.030196444855795965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,256,0.007804444266690149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,1024,0.02311555544535319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,64,0.004618666652176115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,128,0.009773333039548662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,32,0.004943111290534337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,128,0.007299555672539606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,1024,0.013810666898886362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,64,0.008643555144468943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,32,0.01071733319097095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,65536,0.35003111097547746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,65536,0.17694400416480172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,51200,0.27389955520629883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,16384,0.083679113123152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,128,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,51200,0.13442577256096735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,12288,0.0633102191819085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,2560,0.02641155487961239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,10240,0.05369333426157633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,12288,0.0373582210805681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,12288,0.03371289041307237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,10240,0.032277332411872015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,8192,0.044567998912599355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,10240,0.028650667932298448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,7168,0.03878488805558946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,8192,0.024883555041419134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,8192,0.0277804434299469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,51200,0.11823733647664388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,6144,0.033933334880405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,7168,0.02233155568440755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,7168,0.024420445164044697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,5120,0.02940533227390713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,6144,0.019691555036438834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,6144,0.022081777453422546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,5120,0.021410667233996924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,5120,0.01889511115021176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,3584,0.02180800007449256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,65536,0.150691548983256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,256,0.009714666340086196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,3072,0.019271110494931538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,3584,0.01460533340771993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,3072,0.013192888763215808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,2560,0.016548444827397663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,4096,0.024876443876160517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,2560,0.012089778151777057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,4096,0.015809777710172866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,2560,0.012647999657524956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,2048,0.010820444259378644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,3584,0.01480888823668162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,1536,0.011074666347768573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,1536,0.009359999663299983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,1024,0.00868088917599784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,1024,0.007984889050324758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,4096,0.015933333171738517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,768,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,1024,0.008011555506123437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,2048,0.011026666396194033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,768,0.007687110867765214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,512,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,768,0.006680000159475539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,256,0.0038764443662431505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,512,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,1536,0.009692444569534725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,512,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,256,0.00666222224632899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,256,0.00565155554148886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,64,0.0032302221904198327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,128,0.007060444189442529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,128,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,32,0.003195555466744635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,64,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,32,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,3072,0.013111111190583972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,128,0.0035146665241983202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,65536,0.1555724408891466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,2048,0.013457777599493662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,65536,0.1367271078957452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,16384,0.07651999923917982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,16384,0.04177244504292806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,65536,0.2742195659213596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,12288,0.052990221314960055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,16384,0.0373911095990075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,12288,0.03286577926741706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,12288,0.032680888970692955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,51200,0.21239644951290557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,8192,0.037046220567491316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,51200,0.11668622493743896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,10240,0.028338667419221666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,16384,0.04928266670968798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,7168,0.032072888480292425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,8192,0.02166666587193807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,8192,0.024687111377716064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,6144,0.028075555960337322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,7168,0.02015644477473365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,7168,0.02160444524553087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,6144,0.01792622274822659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,6144,0.019275556008021038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,10240,0.043749332427978516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,4096,0.020621332857343886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,5120,0.01779733267095354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,5120,0.01646844380431705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,3584,0.0181031111213896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,4096,0.014304000470373364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,10240,0.025891555680169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,3072,0.016151999433835346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,51200,0.10490755240122478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,3584,0.013289777768982781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,2560,0.013439111411571503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,3072,0.012475555141766867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,2048,0.011418666276666852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,16384,0.04240266813172234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,5120,0.024137778414620295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,1536,0.009667555491129557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,2048,0.010056888891590966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,2048,0.010400888820489248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,1024,0.007861333588759104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,1536,0.008822222550710043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,3584,0.013160000244776407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,4096,0.014411555396185981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,768,0.00685511115524504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,768,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,3072,0.01223111069864697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,768,0.0063999998900625445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,512,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,256,0.0038942222793896994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,1536,0.00868088917599784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,256,0.006642666541867786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,128,0.0032871110985676446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,1024,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,128,0.006999111009968652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,64,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,2560,0.011158222125636207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,32,0.0035911111368073356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,512,0.005980444451173146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,32,0.006618666566080517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,65536,0.21918133894602457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,256,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,1024,0.007017778025733099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,51200,0.17376355330149332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,65536,0.1414168940650092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,128,0.0052764444715446895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,16384,0.06336355871624418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,51200,0.11071999867757161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,51200,0.09909155633714463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,16384,0.03618844350179037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,64,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,16384,0.040250665611690946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,10240,0.03526399864090814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,12288,0.0285173323419359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,12288,0.031086222993002996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,10240,0.02458133300145467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,10240,0.027004443936877783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,7168,0.025892444782786902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,8192,0.021018667353524104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,8192,0.023131555981106226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,2560,0.011409777734014722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,6144,0.022509333160188463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,7168,0.01920355525281694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,5120,0.01965777741538154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,6144,0.01760088900725047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,6144,0.018106665876176622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,4096,0.01683022247420417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,5120,0.01585244470172458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,5120,0.016116445263226826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,3584,0.014536889062987434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,4096,0.014142221874660917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,4096,0.01390044391155243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,3072,0.01316088851955202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,3584,0.01314933349688848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,3584,0.013103110922707452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,512,0.005320888840489917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,2560,0.011163555913501315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,3072,0.012096888489193387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,12288,0.04260799951023526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,2048,0.009910222556855943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,2560,0.011056889262464313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,2048,0.009370666411187913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,1536,0.008368000388145447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,1536,0.008295999632941352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,65536,0.12772888607449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,1024,0.006948444578382704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,7168,0.020277332928445604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,1536,0.008375110725561777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,768,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,2560,0.01069155583779017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,1024,0.006657777975002925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,768,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,8192,0.030208888981077407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,2048,0.010032888915803697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,256,0.003909333298603694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,512,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,512,0.005716444303592046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,128,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,3072,0.012123555772834353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,256,0.005432888865470886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,64,0.003256888853179084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,128,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,32,0.003544888976547453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,64,0.00566133318675889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,768,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,32,0.006065777606434292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,65536,0.17904178301493326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,51200,0.1415297720167372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,65536,0.08993866708543564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,16384,0.04909155435032315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,51200,0.05725600322087606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,128,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,1024,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,12288,0.03669333457946777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,16384,0.024396444360415142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,16384,0.027965333726671007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,10240,0.030573334958818223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,12288,0.018699554933442008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,8192,0.02641244398223029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,10240,0.016530666086408827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,10240,0.01924711134698656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,256,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,7168,0.022965333527988855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,65536,0.07197422451443143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,6144,0.02016177773475647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,7168,0.013665777941544851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,7168,0.014688889185587565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,5120,0.017298666967286002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,6144,0.013449778159459433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,512,0.004234666625658671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,4096,0.014840000205569796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,5120,0.011727111207114326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,12288,0.021965333157115515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,4096,0.010773333410422007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,8192,0.014800000521871777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,3584,0.009713778065310584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,3584,0.010263111028406356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,3072,0.011757333245542316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,6144,0.012858666479587555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,2560,0.010382222632567087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,3072,0.009452444811662039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,5120,0.012067555553383298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,2048,0.008993777963850234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,2560,0.008934222161769867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,3584,0.013055110971132914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,1536,0.007727999654081132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,2048,0.007920888562997183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,1536,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,1536,0.006637333167923822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,1024,0.00591555568906996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,3072,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,768,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,1024,0.006660444455014334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,1024,0.005963555640644497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,512,0.003887999802827835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,768,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,8192,0.01648000048266517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,256,0.003555555724435382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,512,0.005311111195219888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,256,0.005967999911970562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,128,0.0032088889016045462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,128,0.004964444372389051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,2560,0.008345777789751688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,64,0.0032035555276605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,4096,0.010794666906197866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,128,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,32,0.0032088889016045462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,512,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,65536,0.16089955965677896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,64,0.005274666680230035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,32,0.005687110953860813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,51200,0.0721671117676629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,256,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,65536,0.08586577574412028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,16384,0.0469742218653361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,51200,0.054682665401034884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,16384,0.022280888424979314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,768,0.006454222318198945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,12288,0.034726222356160484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,2048,0.007680888805124495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,12288,0.01795377830664317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,12288,0.020262221495310467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,10240,0.015830222103330824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,10240,0.017602667212486267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,8192,0.024259555670950148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,65536,0.06727200084262423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,8192,0.014392000105645923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,8192,0.015509333875444202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,16384,0.026712000370025635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,6144,0.018790221876568265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,7168,0.013163555827405719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,7168,0.013833777772055732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,10240,0.028752889898088243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,6144,0.012356444365448423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,6144,0.012771555946932899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,4096,0.014132445057233175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,5120,0.011055111057228513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,51200,0.12734133667416042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,3584,0.012488889197508493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,4096,0.010447111394670276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,3584,0.008882666627566019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,3072,0.011378666592968835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,5120,0.016280000408490498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,2560,0.010067555639478896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,3072,0.008770667016506195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,5120,0.011412444214026133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,2560,0.007970666719807519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,2560,0.008470222353935242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,7168,0.021306667062971327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,1536,0.007663999994595845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,2048,0.007368000017272101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,2048,0.007003555695215861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,1024,0.005620444400442972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,1536,0.007019555403126611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,1536,0.006362666686375936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,768,0.004554666578769684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,1024,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,768,0.006405333264006509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,768,0.00562755556570159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,51200,0.06564799944559734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,3584,0.009820444716347588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,256,0.003544888976547453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,1024,0.006315555423498154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,3072,0.00906399968597624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,128,0.003206222214632564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,256,0.006000888844331105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,2048,0.008752000000741746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,64,0.003206222214632564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,128,0.004937777916590373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,32,0.0032142222755485107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,512,0.006292444550328785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,64,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,65536,0.1405644416809082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,32,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,51200,0.11197066307067871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,256,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,65536,0.0779155559009976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,65536,0.06478577852249146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,16384,0.03975733452373081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,4096,0.009900444083743626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,51200,0.05947555436028374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,16384,0.02102311121092902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,16384,0.02404533326625824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,512,0.005315555466545953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,12288,0.01754399968518151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,12288,0.018253333038753934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,8192,0.02513066596455044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,10240,0.015558222929636637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,10240,0.01620444489849938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,7168,0.022016889519161646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,8192,0.013787555197874704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,8192,0.014632888966136508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,6144,0.017097777790493436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,12288,0.03267022305064731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,6144,0.011818666425016193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,6144,0.012076444096035428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,5120,0.015241778559154935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,5120,0.010856000085671743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,4096,0.013751111096805997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,512,0.004210666649871402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,7168,0.013154666456911298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,3584,0.011443555355072021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,10240,0.028160889943440754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,3584,0.0086986662613021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,3584,0.009431111315886179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,3072,0.010510221951537661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,5120,0.010958222051461538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,2560,0.009341333475377824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,3072,0.008886222210195329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,128,0.005969777703285217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,4096,0.010034666293197209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,2048,0.007448888487286038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,2560,0.007531555162535773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,4096,0.009177777502271865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,2048,0.007337777978844113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,2048,0.006643555644485686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,1536,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,1536,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,1024,0.00463377767139011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,1024,0.0063484443558586975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,3072,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,768,0.004271111140648524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,7168,0.012929778132173749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,768,0.0063368889192740125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,2560,0.007696000238259633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,768,0.005625777774386936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,1536,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,512,0.006212444355090459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,512,0.0052800000541739995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,128,0.002963555562827322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,256,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,128,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,128,0.005104889058404499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,1024,0.005588444570700328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,32,0.0028773333049482773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,51200,0.05270666546291775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,512,0.003871110992299186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,64,0.0053324442770746015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,256,0.0032088889016045462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,32,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,51200,0.09494577513800727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,65536,0.062301331096225314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,16384,0.035855998595555626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,51200,0.050456000698937305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,64,0.0028657778683635923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,16384,0.020278222031063504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,12288,0.028354667954974707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,16384,0.021807110971874658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,10240,0.024009777439965144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,12288,0.016851555969980028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,12288,0.016869333055284288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,8192,0.02040177749262916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,10240,0.015203555425008139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,7168,0.018749333090252347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,256,0.005975111077229182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,8192,0.013789333403110504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,51200,0.05389599998792013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,6144,0.015799999237060547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,7168,0.012416000167528788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,6144,0.011405333048767514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,6144,0.011437333292431302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,5120,0.010088889135254754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,5120,0.010218666659461128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,10240,0.015153777268197803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,8192,0.013452444639470843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,65536,0.07133155398898654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,3584,0.010704889065689512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,7168,0.01256533298227522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,3072,0.010026666853162978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,3584,0.008887111312813228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,5120,0.014158222410413953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,2560,0.008781332936551835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,3072,0.007974222302436829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,3072,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,65536,0.11850044462415908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,2048,0.007148444652557373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,2560,0.007633777956167857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,2560,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,2048,0.007339555356237624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,4096,0.008692444198661381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,2048,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,4096,0.009369778136412302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,1536,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,1536,0.005969777703285217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,3584,0.00835111074977451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,1024,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,768,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,768,0.005629333357016246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,512,0.0061048890153567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,512,0.0052871112194326185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,1536,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,256,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,256,0.0052764444715446895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,768,0.0042053332759274375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,1024,0.006635555376609166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,64,0.0031964443624019623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,128,0.005968889014588461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,512,0.0039022221333450745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,64,0.005219555563396878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,4096,0.012445333103338877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,256,0.0032257777121331957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,1024,0.004312889029582341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,65536,0.0601466695467631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,51200,0.07333866755167644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,51200,0.048357334401872426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,32,0.0029066666546795103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,16384,0.029096888171301946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,16384,0.01976355579164293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,32,0.005312888986534542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,12288,0.023025777604844835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,16384,0.019539554913838703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,10240,0.020276443825827706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,12288,0.016215110818545025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,128,0.003195555466744635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,12288,0.015781333049138386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,8192,0.01738933391041226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,10240,0.014519999424616495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,128,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,8192,0.01312444441848331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,7168,0.01586577792962392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,7168,0.011728888584507836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,65536,0.09599377711613973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,65536,0.06439022223154704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,6144,0.010750222537252637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,5120,0.011408888631396823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,6144,0.010608000059922537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,51200,0.049916442897584706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,8192,0.013037333057986366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,4096,0.0103511114915212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,5120,0.009726222190592024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,3584,0.00902133352226681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,4096,0.008384000096056197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,6144,0.012671999633312225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,3072,0.008261333737108443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,3584,0.008050666915045844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,3584,0.00794488853878445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,2560,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,3072,0.007681777907742395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,3072,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,2048,0.006320000108745363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,2560,0.007487110793590546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,2560,0.006646222124497096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,1536,0.005144888742102517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,2048,0.00703999979628457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,2048,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,1024,0.004219555606444676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,1536,0.006675555474228329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,768,0.0038693332009845306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,1024,0.006341333190600078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,768,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,512,0.003590222034189436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,10240,0.014144889182514615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,5120,0.009309333231714036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,256,0.0032035555276605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,512,0.0053084443012873335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,1536,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,128,0.0031591111587153543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,1024,0.005630222045713001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,128,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,768,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,64,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,32,0.0028880000528362063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,512,0.005976888868543837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,64,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,32,0.0053431110249625305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,256,0.005978666659858491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,256,0.005088888936572605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,51200,0.06419466601477729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,65536,0.059135999944474965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,16384,0.026573333475324843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,51200,0.04755999975734287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,16384,0.019652444455358718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,16384,0.018556444181336295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,12288,0.021160889003011916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,65536,0.0790168907907274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,10240,0.018431999617152743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,7168,0.01145155562294854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,65536,0.06118844615088569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,10240,0.013760888742076026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,128,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,8192,0.012655110822783576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,7168,0.013331555657916598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,12288,0.01610577768749661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,4096,0.008851555486520132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,7168,0.011083555718262991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,6144,0.01202488856183158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,7168,0.011456888582971362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,5120,0.010686222049925061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,6144,0.009766221874290043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,6144,0.010393777655230628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,5120,0.00905688852071762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,5120,0.009619555539555019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,8192,0.012472888661755456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,3584,0.00869511150651508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,4096,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,4096,0.008218666745556725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,3072,0.008077333370844523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,3584,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,12288,0.0151555554734336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,2560,0.007138667007287343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,3072,0.007777777810891469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,10240,0.014460444450378418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,8192,0.016520889268981088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,51200,0.04797777864668104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,4096,0.009584888815879822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,1536,0.004994666824738185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,2048,0.007296889192528195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,1536,0.006744888921578725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,1024,0.004230222354332606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,3584,0.007675555845101674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,1024,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,1024,0.0063324446479479475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,768,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,2048,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,512,0.0035662220584021676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,2560,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,768,0.00630044440428416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,768,0.005333333379692501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,2048,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,512,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,512,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,128,0.003624888757864634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,256,0.005707555760939916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,256,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,64,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,128,0.004938666605287128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,3072,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,2560,0.007363555332024892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,64,0.0052355556852287715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,65536,0.06979288657506307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,32,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,256,0.003212444484233856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,65536,0.05823644664552477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,65536,0.058881779511769615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,1536,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,51200,0.04625066783693102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,128,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,12288,0.01871288816134135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,16384,0.019291554888089497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,12288,0.016063110695944894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,12288,0.014799111419253878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,51200,0.05611999829610189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,10240,0.013445333474212222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,16384,0.022375999225510493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,8192,0.014170666535695394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,51200,0.04657422171698677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,7168,0.013502221968438892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,16384,0.01814133259985182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,7168,0.010775110787815519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,6144,0.01239911135700014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,10240,0.016384000579516094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,6144,0.009914666414260864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,5120,0.010326222413116032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,6144,0.010191111101044549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,4096,0.009317333499590555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,5120,0.009101333717505137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,10240,0.014096889230940076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,5120,0.008911111288600499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,3584,0.008295111358165741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,4096,0.008375110725561777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,4096,0.007689778175618913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,3072,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,3584,0.007985778152942657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,3584,0.007276444799370236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,3072,0.007705777883529663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,7168,0.011067555182509951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,3072,0.0070017774899800616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,2048,0.005376889059940974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,2560,0.007371555599901412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,2560,0.00658044425977601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,2048,0.0069511110583941145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,2048,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,1024,0.0042462220622433555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,1536,0.006680888848172293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,1536,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,768,0.003918222255176968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,1024,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,1536,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,512,0.0035377778112888336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,768,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,768,0.005285333428117964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,256,0.0032248888164758682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,512,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,512,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,128,0.0030257778449190986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,256,0.004950222041871813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,128,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,128,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,64,0.0028702221396896574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,32,0.002985777747299936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,8192,0.012087999946541257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,2560,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,64,0.004964444372389051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,32,0.005221333354711533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,51200,0.048283557097117104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,256,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,51200,0.045588443676630654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,51200,0.04418666826354133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,1024,0.005668444352017508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,65536,0.060452441374460854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,16384,0.018979556030697294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,65536,0.05717777543597751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,65536,0.05704800287882487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,12288,0.014507555299335055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,12288,0.015652444627549913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,16384,0.020410666863123577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,8192,0.013165333204799227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,32,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,12288,0.016890666551060148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,7168,0.012485333614879184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,8192,0.011260444091426002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,7168,0.010515555739402771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,10240,0.015032000011867948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,6144,0.011734222372372946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,10240,0.013735111388895245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,10240,0.013110222087966071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,8192,0.012094222009181976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,8192,0.011739555332395764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,5120,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,5120,0.00815555536084705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,7168,0.010703110860453712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,4096,0.008361777497662438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,6144,0.009692444569534725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,16384,0.017698667115635343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,3584,0.007286222444640265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,3584,0.008047111332416534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,4096,0.0090506664580769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,3072,0.0068959999415609576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,3072,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,2560,0.006265777680608962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,6144,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,2048,0.0052675555149714155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,2560,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,2048,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,3072,0.0069466663731469055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,1536,0.006697777658700943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,1024,0.004269333349333869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,4096,0.007632888853549957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,3584,0.007985778152942657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,768,0.003899555653333664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,1024,0.005620444400442972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,2048,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,512,0.003544888976547453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,512,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,512,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,5120,0.010757333702511258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,256,0.003212444484233856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,1024,0.0063235556913746735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,128,0.0031946665710873077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,256,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,256,0.005061333378156026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,128,0.005903111149867375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,128,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,32,0.0029564443975687027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,64,0.004966222163703707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,65536,0.05116088853942024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,32,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,2560,0.007627555893527136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,51200,0.04094488753212823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,65536,0.055664002895355225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,65536,0.053631110323799976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,16384,0.01867555578549703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,51200,0.04495822058783638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,51200,0.04251911242802938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,12288,0.015266666809717814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,16384,0.01848266687658098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,16384,0.017086222767829895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,10240,0.013789333403110504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,12288,0.014471999473041959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,12288,0.014085332552591959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,8192,0.013112000293201871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,10240,0.012631110846996307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,10240,0.012780444489585029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,7168,0.011402666568756104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,8192,0.011060444845093621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,8192,0.011067555182509951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,6144,0.010655110908879174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,7168,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,7168,0.010056888891590966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,5120,0.009973333113723332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,6144,0.009716444545321995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,6144,0.009033777647548253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,1536,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,4096,0.009008888569143083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,5120,0.008100444244013892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,3584,0.007632888853549957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,4096,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,3584,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,3072,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,1536,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,3072,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,2560,0.005757333503829108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,5120,0.00906577789121204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,4096,0.008355555435021719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,2048,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,3584,0.007442666424645319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,768,0.0053102220926019884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,1536,0.004948444250557157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,3072,0.006924444602595435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,1536,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,1536,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,1024,0.0042133331298828125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,2560,0.007648000286685095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,768,0.003579555700222651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,1024,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,768,0.00620977787507905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,2048,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,512,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,1024,0.006333333336644703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,2560,0.006585777633719974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,256,0.00309599998096625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,64,0.0028373334142896864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,128,0.0028924445311228433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,256,0.005085333353943295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,2048,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,64,0.002558222247494592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,128,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,128,0.005061333378156026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,32,0.0026826666047175727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,64,0.005099555684460534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,512,0.005974222388532426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,32,0.004954666727119022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,256,0.0058311112225055695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,65536,0.05207111106978523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,512,0.003531555748648114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,51200,0.044590221510993115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,51200,0.04109333289994134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,16384,0.014903111590279473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,16384,0.0177848885456721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,12288,0.013691555294725629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,768,0.006313777632183499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,65536,0.05535466803444756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,12288,0.013791999883121915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,10240,0.012409778104888068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,65536,0.03987110985649957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,16384,0.016792888442675274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,8192,0.010955555571450127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,7168,0.010397333237859938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,8192,0.010788444843557147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,51200,0.03190577692455716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,8192,0.010036444498433007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,7168,0.010382222632567087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,7168,0.009563555320103964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,5120,0.009153777526484596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,6144,0.008780444661776224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,5120,0.009026666482289633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,10240,0.01220355596807268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,4096,0.0074035558435651995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,12288,0.013441777891582914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,3584,0.008034666379292807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,6144,0.009699555734793345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,3584,0.007281777759393056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,3072,0.006320888797442119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,5120,0.008021333151393467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,3072,0.007641777396202087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,2560,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,768,0.005544888890451855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,4096,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,2560,0.006435555716355641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,3584,0.007975999679830339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,2048,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,4096,0.00833422193924586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,1536,0.004521777646409141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,2048,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,1024,0.0039013334446483185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,1536,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,3072,0.006688000013430913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,1536,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,10240,0.0120666664507654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,1024,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,512,0.003567999849716822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,768,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,512,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,256,0.0031591111587153543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,256,0.004969777746333016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,6144,0.009373333719041612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,128,0.0028826666788922418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,2048,0.0052764444715446895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,768,0.006299555715587404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,64,0.002847111059559716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,128,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,512,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,64,0.004961777892377642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,65536,0.029024001624849107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,32,0.005212444398138259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,768,0.003867555409669876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,51200,0.025175111161337957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,65536,0.05459377831882901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,256,0.00565155554148886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,2560,0.00741511086622874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,16384,0.01276711126168569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,51200,0.04382222228580051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,128,0.0059262220230367445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,16384,0.016197333733240765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,32,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,10240,0.010493333141009012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,12288,0.012859555582205454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,10240,0.012198222180207571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,10240,0.01143200033240848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,8192,0.009694221946928237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,1024,0.0063182223174307085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,7168,0.00871377769443724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,8192,0.011002666420406766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,8192,0.010024888647927178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,51200,0.03974666529231601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,7168,0.009381333159075843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,16384,0.01591111057334476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,5120,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,6144,0.009649777577983009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,6144,0.008656000097592672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,4096,0.007627555893527136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,5120,0.008751110898123847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,5120,0.007979555262459649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,3584,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,65536,0.04955910974078708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,3584,0.008037333687146505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,3584,0.0070568886068132185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,3072,0.006640000061856375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,12288,0.013459555804729462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,3072,0.007660444411966536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,2560,0.005868444426192178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,4096,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,4096,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,2048,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,2560,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,7168,0.01036177741156684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,1536,0.004608889006906086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,2048,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,1536,0.006716444260544247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,1536,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,1024,0.005670222143332164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,2560,0.0063884444534778595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,1024,0.006362666686375936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,12288,0.010557333628336588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,2048,0.0069875551594628235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,512,0.003580444388919406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,768,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,768,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,1024,0.003922666526503033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,512,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,512,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,128,0.0029004443850782183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,256,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,64,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,256,0.004956444518433677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,128,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,32,0.0028435554769304064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,128,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,768,0.003564444267087512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,6144,0.010407999985747868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,768,51200,0.020299555526839364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,32,0.005103999955786599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,65536,0.054380443361070424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,65536,0.04846044381459554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,51200,0.04377333323160807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,3072,0.006680000159475539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,51200,0.039396444956461586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,16384,0.01594666639963786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,16384,0.01516266663869222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,768,10240,0.010050666828950247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,12288,0.013437333206335703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,768,65536,0.025402666793929204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,10240,0.01220266686545478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,10240,0.011083555718262991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,64,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,768,7168,0.008447999755541483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,8192,0.01072799993885888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,768,16384,0.010925333533022137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,8192,0.009912889036867354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,7168,0.010080888867378235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,256,0.0032186667538351486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,768,5120,0.008024888734022776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,6144,0.0094373333785269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,6144,0.008719999757077958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,768,4096,0.007672000262472365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,12288,0.012420444852775998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,5120,0.00811822215716044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,768,3584,0.007000889215204451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,4096,0.007353777686754863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,3584,0.008040000167157915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,768,6144,0.007818666597207388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,768,12288,0.00978488889005449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,7168,0.00941777808798684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,768,2560,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,3072,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,5120,0.008861333131790161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,768,2048,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,4096,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,2048,0.007245333658324347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,768,1536,0.004576000074545543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,3584,0.007326222128338284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,768,3072,0.006301333506902059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,768,1024,0.003871110992299186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,1536,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,1536,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,3072,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,768,768,0.003583111282851961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,1024,0.006346666564544042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,1024,0.005662222289376789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,768,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,768,8192,0.00904088881280687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,768,256,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,512,0.0052933332820733385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,2048,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,768,128,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,256,0.005084444251325396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,128,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,128,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,768,512,0.003531555748648114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,768,32,0.002564444517095884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,2560,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,512,0.005921777751710679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,32,0.004982222285535601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,256,0.005712888720962737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,512,51200,0.015543111496501498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,65536,0.05232355660862393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,768,64,0.002519999941190084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,51200,0.03867822223239475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,512,16384,0.01075111081202825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,768,0.0053075556125905775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,512,65536,0.018903111418088276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,512,12288,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,16384,0.01516711049609714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,65536,0.04810133245256212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,12288,0.012451555165979596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,10240,0.011405333048767514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,51200,0.03875289029545254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,512,8192,0.008841777841250101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,10240,0.012142221960756512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,16384,0.015885333220163982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,8192,0.011064888702498542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,8192,0.009917333722114563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,12288,0.013463111387358772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,512,6144,0.00868888861603207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,7168,0.009405333134863112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,6144,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,6144,0.0086986662613021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,64,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,512,4096,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,5120,0.009001777403884465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,5120,0.008093333078755273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,512,7168,0.009350222018029954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,512,3584,0.0069884442620807225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,4096,0.007276444799370236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,512,3072,0.006268444574541516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,3584,0.007185777856243982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,3072,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,3072,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,2560,0.007271111011505127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,2560,0.007333333293596904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,512,2048,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,4096,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,7168,0.010375111467308467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,512,1536,0.004571555389298333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,512,5120,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,1536,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,512,1024,0.0038684445122877755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,1536,0.006657777975002925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,2560,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,512,768,0.003540444291300244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,1024,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,512,10240,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,768,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,768,0.005281777845488654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,512,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,512,2560,0.005590222362014983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,256,0.005097777893145879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,1024,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,512,128,0.0025840000145965153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,256,0.005924444645643234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,3584,0.008023111356629265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,512,512,0.003272888975010978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,128,0.005923555543025334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,128,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,512,256,0.002867555452717675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,64,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,256,65536,0.015614221493403116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,32,0.004975999808973736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,256,51200,0.015152000718646579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,65536,0.04683466752370199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,65536,0.047225776645872325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,256,16384,0.009645333720578088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,51200,0.03772888912094964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,51200,0.03643644518322415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,256,12288,0.009741333623727163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,16384,0.015847999188635085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,16384,0.0148062225845125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,256,10240,0.009343110852771336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,512,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,12288,0.012415111064910889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,256,8192,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,10240,0.012079111403889127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,10240,0.011208888557222156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,256,7168,0.009089777866999308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,8192,0.01072088877360026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,8192,0.010048888623714447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,256,6144,0.008377778033415476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,7168,0.01014755583471722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,7168,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,512,64,0.002543111021320025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,6144,0.009681777821646797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,6144,0.008843555218643611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,256,4096,0.007322666545708974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,5120,0.00816000004609426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,4096,0.008345777789751688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,256,3584,0.00693866651919153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,2048,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,3584,0.007348444726732042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,256,3072,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,12288,0.013125333521101209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,5120,0.009001777403884465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,256,2560,0.005584888988071018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,3072,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,3072,0.006948444578382704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,256,2048,0.004959110998445087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,2560,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,4096,0.007491555478837755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,2560,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,256,1536,0.004554666578769684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,2048,0.006992888947327931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,2048,0.0063235556913746735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,256,1024,0.0038817777401871155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,256,5120,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,1536,0.006727111008432176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,3584,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,1024,0.006355555521117316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,1024,0.00564177789621883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,256,512,0.0032346666687064697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,768,0.005283555636803309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,512,0.005903999838564131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,256,256,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,2048,0.00701155596309238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,256,0.005819555371999741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,512,32,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,256,128,0.0028862222615215513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,128,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,768,0.006221333311663733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,256,64,0.002559111143151919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,128,0.005084444251325396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,256,32,0.002576888849337896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,128,65536,0.015169777803950839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,512,0.0052977779673205475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,65536,0.04613599843449063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,128,51200,0.014479110638300577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,65536,0.04485422372817993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,128,16384,0.009373333719041612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,51200,0.03763644562827216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,256,0.004899555610285865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,1536,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,128,12288,0.009696889254781935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,16384,0.016000888413853116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,16384,0.014862222803963555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,128,10240,0.008719111482302347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,12288,0.01311911145846049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,128,8192,0.00833155545923445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,10240,0.012130666938092975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,10240,0.011411555939250521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,8192,0.010782221953074137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,64,0.004907555464241239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,32,0.004910222358173794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,128,6144,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,7168,0.01038755559258991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,7168,0.00942311104800966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,256,768,0.003555555724435382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,6144,0.008674666285514832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,5120,0.00889066689544254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,128,4096,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,128,7168,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,8192,0.009709333380063375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,4096,0.007318221860461765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,128,5120,0.008046222229798635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,3584,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,128,3072,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,5120,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,3072,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,128,2560,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,4096,0.008381333616044786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,2560,0.007348444726732042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,128,2048,0.005664888769388199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,3584,0.0081386665503184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,12288,0.012430222498046027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,2048,0.006993778049945831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,3072,0.006674666785531574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,1536,0.005994666781690385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,128,3584,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,128,1024,0.003930666794379552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,1536,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,128,768,0.003536000019974179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,6144,0.009535111486911774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,2560,0.006313777632183499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,128,512,0.003206222214632564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,768,0.005385777602593104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,2048,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,512,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,512,0.0052764444715446895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,256,0.005619555711746216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,128,128,0.0028773333049482773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,256,0.004955555415815777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,51200,0.03556711143917508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,128,64,0.0025519999778932994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,128,0.004934222333961063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,768,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,128,1536,0.004375111311674118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,128,256,0.0028888889484935333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,32,0.004931555440028508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,64,51200,0.014831999937693277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,65536,0.044214222166273326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,64,16384,0.00870488915178511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,51200,0.035657776726616755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,1024,0.006333333336644703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,64,12288,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,1024,0.005623111294375525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,64,10240,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,128,32,0.0027093332674768236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,64,8192,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,16384,0.014888889259762235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,12288,0.012483555409643384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,64,0.004599110947714912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,64,65536,0.016521778371598985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,64,6144,0.006233777850866318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,64,5120,0.0069751110341813825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,7168,0.009401777552233802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,64,4096,0.006344888773229387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,6144,0.00869688888390859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,64,3584,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,5120,0.008014221986134848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,128,0.005690666536490123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,4096,0.007640889121426477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,64,2560,0.006264888991912206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,3584,0.006955555743641323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,64,7168,0.006747555401590135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,3072,0.0069751110341813825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,64,1536,0.004262222184075249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,8192,0.009699555734793345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,64,1024,0.0038977778620190094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,2048,0.006048000107208888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,64,768,0.00351200004418691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,1536,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,64,512,0.0031973332580592898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,1024,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,64,3072,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,768,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,64,128,0.0028515555378463534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,64,64,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,64,2048,0.005921777751710679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,256,0.004952888935804367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,128,0.004993777722120285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,32,65536,0.01588444411754608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,2560,0.00664533343580034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,32,51200,0.01517511076397366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,32,0.004610666798220741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,32,16384,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,65536,0.04424622323777941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,32,12288,0.008363555702898238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,51200,0.03566844595803155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,32,10240,0.00738666703303655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,16384,0.014748444159825643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,512,0.0052826665341854095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,12288,0.012426666915416718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,64,32,0.002672000063790215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,32,7168,0.006790222393141852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,10240,0.011095999843544431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,10240,0.011389333340856763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,32,5120,0.007002666592597961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,64,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,32,4096,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,7168,0.009359999663299983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,6144,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,5120,0.008051555189821456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,32,3072,0.006324444380071428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,4096,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,32,2560,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,32,8192,0.006760889043410619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,3072,0.00700800038046307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,2560,0.006351999938488007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,32,6144,0.0063279999627007385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,32,1536,0.004560889055331548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,8192,0.009741333623727163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,2048,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,32,768,0.003583111282851961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,32,3584,0.005923555543025334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,1536,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,32,512,0.0032142222755485107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,1024,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,64,256,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,768,0.005608888963858287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,32,128,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,512,0.005335999859703912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,256,0.004994666824738185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,32,2048,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,128,0.004982222285535601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,64,0.004957333207130432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,32,1024,0.003920000046491623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,32,0.004597333156400257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,32,256,0.002847111059559716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,3584,0.006999111009968652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,51200,0.49431822035047746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,16384,0.36600976520114475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,32,32,0.002516444358560774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,51200,1.112601810031467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,16384,0.16642844676971436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,12288,0.2734791172875298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,51200,0.42452356550428605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,12288,0.12416444884406196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,49,32,64,0.002514666567246119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,10240,0.22951555252075195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,10240,0.10514844126171535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,8192,0.18325866593254936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,10240,0.08593422174453735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,12288,0.2125413285361396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,16384,0.1491182247797648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,8192,0.08879199955198501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,7168,0.1611831055747138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,12288,0.10231733322143555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,8192,0.07015022304322985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,10240,0.17718044916788736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,6144,0.13899200492435032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,7168,0.07526755332946777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,8192,0.14794311258527967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,6144,0.0659431086646186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,16384,0.28202933735317653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,5120,0.12039555443657769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,6144,0.054692443874147206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,6144,0.11231288645002578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,5120,0.05554577708244324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,4096,0.09497066338857015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,7168,0.07304088936911689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,5120,0.09222666422526042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,4096,0.04628888765970866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,3584,0.0841368900405036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,4096,0.03937955697377523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,7168,0.13118844562106663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,3584,0.04132711225085788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,3584,0.0403386652469635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,3584,0.06906044483184814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,5120,0.04668800036112467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,2560,0.061716443962521024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,3072,0.03147466646300422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,3072,0.06154400110244751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,51200,0.8980435265435113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,3072,0.03629866573545668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,2560,0.03209066722128127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,2560,0.027056000298923914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,2560,0.0582853356997172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,4096,0.07742844687567817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,3072,0.07451555464002821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,2048,0.04774577750100029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,1536,0.018836443622907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,1536,0.03901422354910109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,1024,0.016403555870056152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,2048,0.0503422220547994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,1024,0.015272888872358533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,1024,0.031545778115590416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,768,0.022111111217074927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,2048,0.027403554982609216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,512,0.015235554840829639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,768,0.0138568886452251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,1536,0.02201777696609497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,1024,0.02923733327123854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,256,0.00924355536699295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,512,0.012305777933862476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,512,0.02258133391539256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,256,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,256,0.01074577785200543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,128,0.0075733330514695905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,1536,0.03903822104136149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,2048,0.023175110419591267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,128,0.010880889164076911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,128,0.015784889459609985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,32,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,64,0.009703999592198266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,512,0.010743111371994019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,32,0.01277955538696713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,256,0.020236444142129686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,128,0.008341333104504479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,65536,0.5232444339328343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,768,0.02685244381427765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,65536,1.1112835142347548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,51200,0.3974933359358046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,51200,0.8653670946756998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,64,0.006959111326270633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,16384,0.28298844231499565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,768,0.013452444639470843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,12288,0.21332177850935194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,16384,0.11573510699801975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,12288,0.10008355644014146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,65536,0.42837243609958225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,10240,0.17558666070302328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,12288,0.08505511283874512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,16384,0.22716177834404838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,16384,0.13346844249301487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,51200,0.3272435665130615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,10240,0.08497866657045151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,8192,0.14292889171176487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,10240,0.07278577486673991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,7168,0.12874933083852133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,8192,0.07135288582907783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,8192,0.11836177772945827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,12288,0.17329155074225533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,7168,0.059319999482896596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,6144,0.10923466417524551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,8192,0.05707644753985935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,6144,0.05369155605634054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,5120,0.09209333525763617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,6144,0.044377777311537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,7168,0.06122044722239176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,5120,0.04558222161398994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,5120,0.04098133246103922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,6144,0.09252799881829156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,5120,0.07864089144600762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,4096,0.0378479990694258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,3584,0.0667768915494283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,4096,0.06607022550370958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,10240,0.14612800545162624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,65536,0.957906617058648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,3584,0.03418933351834615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,7168,0.10602311293284099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,3072,0.05694755580690172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,3072,0.03039466672473484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,3072,0.05296088920699226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,3072,0.02681955529583825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,2560,0.04854399959246317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,3584,0.03477777706252204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,4096,0.07468977901670668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,2560,0.026410667432679072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,4096,0.0329751107427809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,2048,0.02294755478700002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,2048,0.019259555472267997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,1536,0.031594667169782854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,2560,0.026495110657480028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,3584,0.059692442417144775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,2048,0.04003377755482992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,1536,0.01611288885275523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,51200,0.727092425028483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,1024,0.013821333646774292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,1024,0.013431111143694984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,2048,0.04037422272894118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,1024,0.027560888065232172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,768,0.012171555724408893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,512,0.011993777420785693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,2560,0.046642667717403836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,1024,0.02279022170437707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,1536,0.033891555335786604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,256,0.007982222570313348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,512,0.010715555813577441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,512,0.01982399986849891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,768,0.016873778568373788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,256,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,256,0.009692444569534725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,256,0.017879999346203275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,768,0.012288000020715924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,128,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,128,0.009696000152164036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,128,0.014141332772043018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,32,0.004224888980388641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,64,0.00868711123863856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,512,0.009964444571071202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,32,0.010779555473062726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,65536,0.34986223114861387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,128,0.004934222333961063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,51200,0.2733786635928684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,65536,0.15153777599334717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,64,0.004190222256713443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,768,0.023448000351587932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,51200,0.13432355721791586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,51200,0.1207502285639445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,16384,0.044460445642471313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,65536,0.17892355389065215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,65536,0.17961511347028944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,16384,0.08257066541247897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,51200,0.13803999953799778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,10240,0.053724444574779935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,16384,0.04898311032189263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,12288,0.04135200050142076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,16384,0.05453422334459093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,12288,0.06432000133726332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,10240,0.03584355447027419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,10240,0.02832533253563775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,1536,0.018430221411916945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,8192,0.024078221784697637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,8192,0.03153511219554477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,12288,0.033667554457982383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,7168,0.024336000283559162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,6144,0.03390311201413473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,7168,0.02567022211021847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,7168,0.028006222512986925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,8192,0.04518844352828132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,6144,0.019855111837387085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,6144,0.021941334009170532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,5120,0.029366221692827012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,6144,0.025412445267041523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,7168,0.03862133291032579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,5120,0.018887110882335238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,5120,0.022456000248591106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,4096,0.015895111693276297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,4096,0.015531554818153381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,3584,0.021529777182473078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,3584,0.014692443940374585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,3584,0.018279999494552612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,3072,0.018943111101786297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,3584,0.01588088936275906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,5120,0.017435555656750996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,3072,0.01311822235584259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,2560,0.01629066632853614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,3072,0.016883555385801528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,4096,0.01956533392270406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,2560,0.012080888781282636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,2560,0.012661333713266583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,2048,0.013086222112178802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,2560,0.015536889433860779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,8192,0.02813777658674452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,2048,0.011025778121418424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,2048,0.01163999984661738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,2048,0.01424711114830441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,3072,0.013251555462678274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,1536,0.009685333404276106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,10240,0.032235556178622775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,1536,0.01310400002532535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,1024,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,1024,0.0114720000161065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,12288,0.03734488950835334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,1536,0.0110871113008923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,768,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,1536,0.009773333039548662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,768,0.009359999663299983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,512,0.005640000104904175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,768,0.007666666474607255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,512,0.007465778125656976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,512,0.006015111174848344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,256,0.0038728887836138406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,4096,0.025111999776628282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,256,0.005595555735958948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,128,0.003228444399105178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,256,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,1024,0.008661333057615492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,128,0.0053635554181204895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,64,0.0032008888406885993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,512,0.00871733327706655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,32,0.0032319999817344877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,768,0.007270221908887227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,256,0.006735111276308696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,65536,0.25971465640597874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,65536,0.153111113442315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,128,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,128,0.007710222568776872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,51200,0.20631110668182373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,65536,0.1294355524910821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,16384,0.07427199681599934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,51200,0.11635022693210179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,1024,0.008056888977686564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,51200,0.10151555803087021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,32,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,12288,0.05317066775427925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,16384,0.04061866799990336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,16384,0.04482400086190966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,65536,0.14292444123162165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,10240,0.04354222284422981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,12288,0.031698667340808444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,51200,0.11284532811906602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,8192,0.03666933377583822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,10240,0.0258577780591117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,64,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,10240,0.03108355402946472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,7168,0.03208977646297879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,8192,0.024596444434589807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,8192,0.027506666051016912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,16384,0.04265244470702278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,7168,0.021533333592944678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,6144,0.028074665202034846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,12288,0.03261244297027588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,7168,0.02247200078434414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,5120,0.024096889628304377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,6144,0.01811288959450192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,6144,0.019303111566437613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,5120,0.016197333733240765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,8192,0.02442666722668542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,4096,0.02127555509408315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,6144,0.022346667117542688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,5120,0.016533333394262526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,3584,0.0181057784292433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,4096,0.01442399952146742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,4096,0.014561777313550314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,4096,0.017655111021465726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,7168,0.024199111594094172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,3072,0.01587288909488254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,3584,0.014496889379289416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,3584,0.013431111143694984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,12288,0.0348417792055342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,3072,0.012489777472284106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,3072,0.015547555353906421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,2560,0.011492444409264458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,2048,0.012390221986505719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,2048,0.010558221903112201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,10240,0.02831733226776123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,3072,0.01218844453493754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,1536,0.010094222095277574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,5120,0.019889778561062284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,2560,0.014524444937705994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,1536,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,1536,0.012430222498046027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,2560,0.013879111243618859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,1024,0.007328000333574083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,1024,0.007641777396202087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,2048,0.013827555709415011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,2560,0.012143111063374413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,1024,0.010107555323176915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,768,0.007359111474619971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,1536,0.008982222113344405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,768,0.009360888765917884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,512,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,512,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,512,0.006993778049945831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,256,0.0038844446341196695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,2048,0.00979377743270662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,768,0.006675555474228329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,128,0.0035413333939181436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,256,0.005625777774386936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,256,0.007991111112965478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,1024,0.008013333711359236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,128,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,128,0.00666933341158761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,3584,0.016604445046848722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,128,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,32,0.003564444267087512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,64,0.005982222242487802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,65536,0.218086216184828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,32,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,256,0.006614222294754452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,512,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,65536,0.1421253283818563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,51200,0.17226844363742402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,64,0.0032097777972618737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,51200,0.10998577541775173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,16384,0.06282489167319404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,51200,0.097853336069319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,12288,0.04233155647913615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,16384,0.03762400150299072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,65536,0.13519377178615993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,16384,0.04198399848408169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,10240,0.0350542234049903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,12288,0.03104355600145128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,12288,0.02940533227390713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,65536,0.12354844146304661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,8192,0.03012266755104065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,10240,0.024640889631377325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,10240,0.029158221350775823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,8192,0.023238221804300945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,7168,0.026007999976476032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,16384,0.040236443281173706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,8192,0.025570667452282373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,51200,0.10759466886520386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,7168,0.020219556159443326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,12288,0.03310488992267185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,7168,0.0233324451578988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,10240,0.026995556222067937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,6144,0.018254222141371835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,5120,0.020064888728989493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,6144,0.021394666698243883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,8192,0.0228426671690411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,5120,0.015986666083335876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,4096,0.017179555363125272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,5120,0.018923555811246235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,768,0.006694222076071634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,4096,0.013667555318938361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,3584,0.015018666783968607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,4096,0.017260443833139207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,3584,0.012900444368521372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,6144,0.01759377784199185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,3072,0.013470222552617392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,3584,0.016591111818949383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,3072,0.012076444096035428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,5120,0.016148444679048326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,2560,0.013440888788965015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,3072,0.015085332923465304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,7168,0.019311111834314134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,6144,0.022825777530670166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,2048,0.011565333439244164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,2560,0.014441778262456259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,3584,0.014139556222491793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,2048,0.013097777962684631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,1536,0.009662222531106737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,2048,0.01015377789735794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,3072,0.01183999992079205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,1536,0.008377778033415476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,1536,0.008352888955010308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,1536,0.011669333610269757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,2560,0.011369778050316704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,2560,0.010785777535703448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,1024,0.007637333538797166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,1024,0.0069884442620807225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,1024,0.009445333646403419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,512,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,768,0.007272889216740926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,768,0.008709333009190029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,2048,0.009684444301658207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,512,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,512,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,512,0.008625778059164682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,1024,0.00798044436507755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,128,0.0035457776652442086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,256,0.006646222124497096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,4096,0.014121777481502958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,768,0.006249777972698212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,64,0.0032515554792351196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,128,0.0052764444715446895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,32,0.0035333335399627686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,128,0.007977777885066139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,64,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,65536,0.1767431100209554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,256,0.004325333154863781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,65536,0.09406666623221503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,256,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,65536,0.07079288694593641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,51200,0.14295821719699436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,128,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,51200,0.07173066669040255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,16384,0.0491484436723921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,51200,0.057349330849117704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,65536,0.13459289073944092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,32,0.006178666734033161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,12288,0.03714222378200955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,16384,0.023152889476882085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,16384,0.040970666540993586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,256,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,12288,0.018831999765502084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,10240,0.03110755483309428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,12288,0.03264710969395108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,768,0.006311111152172089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,8192,0.026534222894244727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,10240,0.01923377811908722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,10240,0.016565332810084026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,10240,0.028816001282797918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,7168,0.023039110832744177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,8192,0.015017777681350708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,8192,0.016569778323173523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,8192,0.025011556016074285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,6144,0.01994755533006456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,7168,0.013995556367768182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,12288,0.0218213333023919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,7168,0.022816888160175745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,6144,0.0127688886390792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,6144,0.013440888788965015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,6144,0.02096533278624217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,4096,0.01497155593501197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,5120,0.011737777955002256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,16384,0.02814133299721612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,4096,0.010913777682516309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,3584,0.013122667041089801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,4096,0.017151999804708693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,7168,0.014713777436150445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,3584,0.010083555347389644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,5120,0.017529777354664274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,3072,0.011733333269755045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,3584,0.016136889656384785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,3072,0.009007111191749573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,3072,0.009541333549552495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,2560,0.010133333504199982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,3072,0.014781332678265043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,2560,0.008656000097592672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,2560,0.009062222308582729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,2048,0.008665777742862701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,2560,0.01408622165520986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,3584,0.01038577738735411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,2048,0.007863110966152614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,1536,0.007668444679843054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,2048,0.01276266657643848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,5120,0.012080888781282636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,1536,0.007155555817815993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,1024,0.005924444645643234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,1536,0.010975110861990186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,1024,0.006737777756320105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,1024,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,1024,0.009451555709044138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,2048,0.007825777762466008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,768,0.008440888590282863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,512,0.0039022221333450745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,1536,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,5120,0.018585777944988675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,512,0.005592888842026393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,512,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,768,0.004548444516128964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,256,0.005230222311284807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,256,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,128,0.0031991110493739447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,768,0.006315555423498154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,256,0.005975111077229182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,128,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,128,0.004937777916590373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,128,0.007970666719807519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,51200,0.10726488961113824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,32,0.0032586666444937387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,64,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,65536,0.15680978033277723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,768,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,32,0.006072888771692912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,65536,0.08584533135096233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,51200,0.1245688862270779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,64,0.0032088889016045462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,51200,0.06536710924572416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,4096,0.010664000279373594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,512,0.00851022203763326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,16384,0.04753244585461087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,51200,0.05452088846100701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,65536,0.13452533880869547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,16384,0.021817778547604878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,16384,0.02628177735540602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,16384,0.04059733284844293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,12288,0.01797066628932953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,12288,0.020270221763186984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,10240,0.01756355497572157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,256,0.003548444559176763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,8192,0.024469332562552557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,12288,0.037215110328462385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,8192,0.016120000018013846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,51200,0.10758044322331746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,10240,0.02884622083769904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,8192,0.014299554957283868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,65536,0.06731644603941175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,7168,0.013797333670987023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,10240,0.028572443458769057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,7168,0.02236533330546485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,6144,0.012469333079126147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,7168,0.021330666210916307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,12288,0.03244444396760728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,6144,0.012237333589129977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,5120,0.016338666280110676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,6144,0.020596444606781006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,10240,0.01590844492117564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,5120,0.01144088887506061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,5120,0.018437332577175565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,4096,0.01036266651418474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,4096,0.01685066686736213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,8192,0.024774221910370722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,3584,0.010036444498433007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,3584,0.015280000037617154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,7168,0.013108444710572561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,3072,0.011024889018800525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,3072,0.009158222211731805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,3584,0.012662221988042196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,3072,0.008428444465001425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,2560,0.008440000315507254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,3584,0.009071110851234859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,2560,0.007962666451931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,2560,0.013781333135233985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,2048,0.008657777474986183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,3072,0.014479110638300577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,2560,0.009930666950013902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,2048,0.007747555772463481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,2048,0.012692444854312472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,1536,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,5120,0.011405333048767514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,1536,0.01072533345884747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,1536,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,4096,0.00977244476477305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,1024,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,768,0.004592000196377437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,1024,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,1536,0.0074924445814556545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,768,0.006308444258239534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,768,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,512,0.003951999876234266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,768,0.009022222624884712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,4096,0.014474666780895658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,1024,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,256,0.003253333270549774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,512,0.005341333233647876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,512,0.008430221842394935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,128,0.0031911111954185697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,256,0.0052977779673205475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,256,0.007982222570313348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,2048,0.0074826669361856245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,64,0.0028773333049482773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,128,0.007634667058785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,32,0.0031928889867332247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,6144,0.01886311173439026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,512,0.006292444550328785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,65536,0.13602311081356472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,32,0.005762666463851929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,256,0.005913777897755305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,65536,0.07759555843141344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,1024,0.009315555294354757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,65536,0.06511555777655707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,128,0.005922666854328579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,16384,0.03896888759401109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,51200,0.05948533614476522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,51200,0.052968889474868774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,65536,0.13389155599806044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,64,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,12288,0.03147555390993754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,51200,0.1066124439239502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,16384,0.040475553936428495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,10240,0.027588443623648748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,51200,0.10838311248355442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,128,0.004982222285535601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,10240,0.01628533336851332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,8192,0.022807111342748005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,10240,0.015772443678643968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,16384,0.021341333786646526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,16384,0.024103111690945093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,7168,0.02035288843843672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,8192,0.01401511165830824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,8192,0.02442488902144962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,12288,0.017515555024147034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,6144,0.015402666396564908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,12288,0.03230311142073737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,7168,0.013049778011110095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,7168,0.021785777476098802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,6144,0.011896889242861005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,6144,0.012104888757069906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,10240,0.028343998723559912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,5120,0.010871111518806882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,8192,0.014521777629852295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,12288,0.018394667241308425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,4096,0.009752000371615091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,4096,0.009301332963837517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,5120,0.013603555659453074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,3584,0.01069777790043089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,6144,0.020422221885787118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,5120,0.010878221856223213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,3584,0.009512888888518015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,3072,0.009682666924264696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,3584,0.015708444847000968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,7168,0.012847111456924014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,3072,0.008800889054934183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,2560,0.008660444782839881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,4096,0.016492444607946608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,2560,0.007628444168302748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,4096,0.01200799975130293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,2048,0.007978666987684038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,3584,0.00869688888390859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,2048,0.006801777829726537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,2048,0.011728000309732227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,1536,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,1536,0.006239111224810283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,5120,0.018266666266653273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,1536,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,1024,0.00490399988161193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,1536,0.010386666489972008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,2048,0.007296889192528195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,1024,0.005614222337802251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,1024,0.009409777820110321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,1024,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,768,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,512,0.003900444342030419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,512,0.005280888742870755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,3072,0.014911111858155994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,768,0.004229333251714706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,256,0.0032497776879204642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,512,0.005987555616431766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,2560,0.013295999831623502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,768,0.009015999734401703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,256,0.005031110925806893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,128,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,256,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,128,0.005963555640644497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,128,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,64,0.0028880000528362063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,128,0.007780444290902879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,3072,0.008103110724025303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,65536,0.11609421836005317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,64,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,32,0.005656888915432825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,2560,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,65536,0.06274488899442884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,256,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,65536,0.1335795587963528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,51200,0.05389777819315592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,16384,0.035226666265063815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,51200,0.05035555693838331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,512,0.007979555262459649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,768,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,12288,0.02734666731622484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,65536,0.07143377595477633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,12288,0.016869333055284288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,32,0.003192000091075897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,12288,0.03201422095298767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,12288,0.01700888905260298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,51200,0.10668444633483887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,16384,0.02195288903183407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,10240,0.028011557128694322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,51200,0.09256800015767415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,8192,0.013510222236315409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,8192,0.023954666323131982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,10240,0.023240889112154644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,7168,0.018183999591403537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,7168,0.012451555165979596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,7168,0.0221413340833452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,8192,0.019859555694792006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,6144,0.016401777664820354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,7168,0.012431999875439538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,6144,0.011416888899273343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,6144,0.011411555939250521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,6144,0.020212444994184706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,8192,0.013433777623706393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,5120,0.010138666464222802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,5120,0.010257778068383535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,5120,0.018188445104493033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,16384,0.020384889509942796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,10240,0.015388444066047668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,4096,0.009004444711738164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,4096,0.016173332929611206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,5120,0.01386222243309021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,3584,0.0087013335691558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,3072,0.010023111270533668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,3584,0.008411555654472774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,3584,0.015279110934999255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,4096,0.012522666818565793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,2560,0.008671999805503422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,3072,0.014018666413095264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,10240,0.01519466605451372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,16384,0.04012177719010247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,2560,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,2560,0.007860444486141205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,2560,0.01276177747382058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,2048,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,2048,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,1536,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,3072,0.008004444340864817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,1536,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,2048,0.00775733341773351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,1536,0.010403555300500657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,1024,0.004563555535342959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,1024,0.005761777775155173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,2048,0.011735999749766456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,3072,0.007640000018808577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,4096,0.009733333355850643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,768,0.0052728888889153796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,1536,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,768,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,768,0.008671110702885522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,512,0.005292444593376583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,1024,0.009015111459626092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,256,0.0035342222286595237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,512,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,3584,0.011062222222487131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,256,0.005078222188684675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,128,0.003198222153716617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,256,0.005990222096443176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,512,0.0035537779331207275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,64,0.0028373334142896864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,128,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,128,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,32,0.0032008888406885993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,128,0.008036444584528605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,1024,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,65536,0.09240711397594875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,64,0.0052728888889153796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,32,0.005642666584915585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,51200,0.0734951098759969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,65536,0.06437955300013225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,51200,0.049569778972201876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,51200,0.0485208895471361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,16384,0.02863022353914049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,51200,0.1065982249048021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,16384,0.01958666741847992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,16384,0.019704888264338177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,12288,0.02291999922858344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,768,0.0041928887367248535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,65536,0.05987022320429484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,12288,0.016418667303191293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,12288,0.03172711200184292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,12288,0.015652444627549913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,256,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,8192,0.01757688820362091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,10240,0.01479288935661316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,10240,0.027469333675172593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,10240,0.01423288881778717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,7168,0.016045333610640634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,8192,0.012860444684823355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,512,0.00832799987660514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,7168,0.011811555259757571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,8192,0.024233778317769367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,6144,0.01349955548842748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,7168,0.02176711128817664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,7168,0.011545778148704104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,65536,0.13357422086927626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,5120,0.01145511120557785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,6144,0.010378667049937779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,6144,0.02014311154683431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,5120,0.009695111049546136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,8192,0.013061333033773633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,16384,0.03990755478541056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,5120,0.018264888061417472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,3584,0.009019555317031013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,4096,0.008835555778609382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,4096,0.008346666892369589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,4096,0.015992888145976596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,3072,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,3584,0.007804444266690149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,3584,0.01480888823668162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,3072,0.007674666742483775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,6144,0.010687111152542962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,3072,0.01387733303838306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,2560,0.0069751110341813825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,10240,0.019976889093716938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,2048,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,2560,0.007351111206743453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,5120,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,4096,0.01036711119943195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,1536,0.005244444641802046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,2048,0.007239110767841339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,1536,0.005979555762476391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,1536,0.01038044442733129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,3584,0.008079999850855933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,1024,0.0042257776690853965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,1024,0.005595555735958948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,2048,0.0062977779242727495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,1024,0.009350222018029954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,768,0.0038613333470291565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,1024,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,768,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,768,0.005584888988071018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,768,0.00869511150651508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,1536,0.006780444333950679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,512,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,256,0.003208000005947219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,512,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,3072,0.007890666524569193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,2048,0.011775111158688864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,256,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,128,0.003002666764789157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,256,0.007705777883529663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,512,0.00360000009338061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,128,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,2560,0.0125608891248703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,128,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,32,0.0029200000895394217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,2560,0.007350222104125553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,64,0.005320000151793162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,65536,0.0791342192225986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,32,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,256,0.005980444451173146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,65536,0.06089511182573107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,65536,0.05901333358552721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,51200,0.06381955411699083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,128,0.004905777672926585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,64,0.002899555489420891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,16384,0.025046222739749487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,512,0.008013333711359236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,51200,0.10623288816875881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,12288,0.020199111766285367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,16384,0.0195448895295461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,16384,0.0398595564895206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,12288,0.015166223049163818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,12288,0.031230221192042034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,65536,0.13361244731479222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,10240,0.014117333624098035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,10240,0.014477334088749357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,10240,0.02739288906256358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,51200,0.047876444127824574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,16384,0.018626666731304593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,8192,0.012476444244384766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,8192,0.012453333371215396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,8192,0.0236435549126731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,12288,0.016207110550668504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,6144,0.013184888495339287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,7168,0.011083555718262991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,7168,0.02163911031352149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,6144,0.010371555884679159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,6144,0.010741333166758219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,5120,0.010823110739390055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,7168,0.014123555686738757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,5120,0.01772622267405192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,5120,0.009083555804358589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,7168,0.011390222443474663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,4096,0.007982222570313348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,51200,0.047863112555609814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,6144,0.020026667250527274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,4096,0.016180444094869826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,5120,0.009728888670603434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,3584,0.007378666765160031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,3072,0.00795911086930169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,4096,0.010033778018421596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,10240,0.01756444407833947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,3072,0.007057777709431118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,2560,0.007326222128338284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,3072,0.014109333356221518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,8192,0.015283554792404175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,3584,0.008626666333940294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,2048,0.006067555397748947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,2560,0.007636444436179266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,2560,0.012103111379676394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,3584,0.014896889527638754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,3584,0.008076444268226624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,1536,0.005151111218664381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,2048,0.011415999796655444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,1536,0.006772444479995304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,1536,0.005964444329341252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,1536,0.010379555324713389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,4096,0.008396444221337637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,1024,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,768,0.003670222229427762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,1024,0.009011555876996782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,2048,0.007117333511511485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,2048,0.006331555545330048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,512,0.003660444584157732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,768,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,768,0.00868000007337994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,3072,0.008003555238246918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,1024,0.004219555606444676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,512,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,512,0.0052986666560173035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,512,0.00833777752187517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,128,0.0028924445311228433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,1024,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,256,0.007636444436179266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,128,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,64,0.0028560000161329904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,768,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,32,0.0029057777590221832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,256,0.0032097777972618737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,64,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,32,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,2560,0.006719111154476802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,65536,0.05880888965394762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,65536,0.057999114195505776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,51200,0.05555911196602715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,65536,0.1332577731874254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,128,0.007958222594526079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,51200,0.04569066564242045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,16384,0.021241777473025854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,65536,0.07031466563542683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,51200,0.10609689023759629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,256,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,12288,0.017717333303557504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,16384,0.0181048893266254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,16384,0.039684444665908813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,12288,0.014575999644067554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,10240,0.01575288838810391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,51200,0.04654666781425476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,256,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,16384,0.01924800044960446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,10240,0.027436445156733196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,8192,0.013788444300492605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,128,0.0059262220230367445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,12288,0.015849777393870883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,8192,0.012440000143316058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,12288,0.031574222776624895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,8192,0.02369511127471924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,7168,0.0111191107167138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,6144,0.011897777517636618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,10240,0.013224889006879596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,10240,0.014183110660976835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,5120,0.01053600013256073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,7168,0.012984889249006907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,5120,0.009073778159088558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,7168,0.011251555548773872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,5120,0.009151111046473185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,5120,0.017854221993022494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,6144,0.010016000105275048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,6144,0.019923556182119582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,3584,0.008375999828179678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,8192,0.01203555530971951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,3584,0.007269333634111617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,4096,0.00941777808798684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,3584,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,4096,0.007680888805124495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,3072,0.006999111009968652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,4096,0.00832177781396442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,3072,0.007749333149856991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,3072,0.013491555220550962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,2560,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,7168,0.02163733376397027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,3072,0.008054222497675154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,2560,0.012458666331238218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,2048,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,2048,0.011408888631396823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,2560,0.007128000259399414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,1536,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,3584,0.014128888646761576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,2048,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,2560,0.007610666255156199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,1536,0.006741333338949416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,1536,0.010159999959998662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,6144,0.00983911090426975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,1024,0.005608888963858287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,768,0.003920000046491623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,1024,0.00907733374171787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,2048,0.006994666324721442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,512,0.003567111161020067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,768,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,768,0.005757333503829108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,768,0.008661333057615492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,256,0.0032266666077905228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,512,0.0052897776994440295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,1024,0.004557333472702238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,512,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,128,0.0028497777465316984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,256,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,256,0.005286222116814719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,4096,0.015392889579137167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,64,0.0028808888875775864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,1024,0.006314666734801398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,32,0.0028853333658642242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,64,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,65536,0.05972799989912245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,32,0.0052782222628593445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,65536,0.0566435522503323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,256,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,128,0.005973333285914526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,65536,0.05685600307252672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,128,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,51200,0.04543022314707438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,512,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,51200,0.1061040030585395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,16384,0.017532444662517972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,16384,0.03943466809060838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,16384,0.01904444396495819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,12288,0.016886222693655226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,1536,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,12288,0.014468444718254937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,51200,0.044285333818859525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,10240,0.015098666151364645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,16384,0.01954844428433312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,10240,0.013450666434235044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,10240,0.013120888835854001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,10240,0.0273964438173506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,51200,0.04703111118740506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,128,0.007618666523032718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,12288,0.015885333220163982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,12288,0.03121688961982727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,8192,0.012530666258600024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,7168,0.01092977739042706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,7168,0.010545777777830759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,6144,0.011120888921949597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,7168,0.02164799968401591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,8192,0.011744889120260874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,6144,0.009345778160625035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,7168,0.012231999801264869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,5120,0.010076444182131026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,5120,0.017543110582563613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,5120,0.009187555147541894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,5120,0.008723555339707268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,8192,0.011730666789743634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,4096,0.00869599978129069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,8192,0.02355466617478265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,4096,0.007747555772463481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,6144,0.009740444521109263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,4096,0.015287111202875773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,3072,0.007480888730949826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,3584,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,4096,0.008986666798591614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,65536,0.13344266679551867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,2560,0.006604444649484422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,3072,0.012862222062216865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,3072,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,2560,0.0074275558193524676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,3584,0.007979555262459649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,2048,0.005919111271699269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,6144,0.01959911154376136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,2048,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,2048,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,2048,0.011404444773991903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,3584,0.014115555418862237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,3072,0.007896000312434303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,1024,0.004253333227501975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,1536,0.006861333217885759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,2560,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,2560,0.012111999922328524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,1024,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,3584,0.008353778057628209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,1024,0.009402666654851701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,768,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,512,0.003544888976547453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,1536,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,1536,0.010453333457310995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,512,0.008325333396593729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,256,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,768,0.0036595554815398324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,1536,0.005250666704442766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,256,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,768,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,256,0.007479111353556315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,768,0.00870666652917862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,128,0.005921777751710679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,64,0.0028373334142896864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,128,0.007615110940403408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,32,0.0028773333049482773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,512,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,1024,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,65536,0.048010667165120445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,32,0.004948444250557157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,256,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,51200,0.03825155562824673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,65536,0.05552800165282356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,65536,0.13333600097232395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,128,0.004963555683692296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,512,0.005250666704442766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,51200,0.04260177744759453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,51200,0.044920888211992055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,51200,0.10560266839133368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,16384,0.018533334136009216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,16384,0.017147555947303772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,16384,0.03928088810708787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,128,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,10240,0.01423377792040507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,12288,0.030982222821977403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,16384,0.017437333861986797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,10240,0.02727288835578495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,10240,0.012688889271683164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,12288,0.01518133282661438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,65536,0.05408177773157755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,12288,0.014815110299322339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,8192,0.023376888699001733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,7168,0.011725333001878528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,8192,0.011418666276666852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,7168,0.010370666782061258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,7168,0.009694221946928237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,6144,0.010709332923094431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,7168,0.021337777376174927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,8192,0.012883555557992725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,64,0.00490133340160052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,6144,0.009863110880057016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,6144,0.01959288948112064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,8192,0.011065777805116443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,5120,0.008040889269775814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,12288,0.014163555370436775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,5120,0.01722755531469981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,10240,0.012753778033786349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,4096,0.0075786668393347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,3584,0.007708444363541073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,4096,0.014854222536087036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,4096,0.008525333470768398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,5120,0.010074666804737514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,6144,0.00906844437122345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,3584,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,3584,0.013963555296262106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,4096,0.00869333330127928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,2560,0.006131555471155379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,3072,0.013156444662147097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,3072,0.006950221955776215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,2560,0.006735999965005451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,2560,0.012549333274364471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,2560,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,5120,0.009022222624884712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,2048,0.006320000108745363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,1536,0.004600888739029567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,1536,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,3584,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,1024,0.0041288890772395665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,1536,0.006682666639486949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,1536,0.010399999717871347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,1024,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,1024,0.005578666511509154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,768,0.003871110992299186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,1024,0.009391999906963771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,768,0.0063697778516345555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,512,0.0032640000184377036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,2048,0.005274666680230035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,512,0.0053635554181204895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,2048,0.007205333146784041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,512,0.006070222291681502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,256,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,512,0.008444444172912175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,768,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,128,0.002846222163902389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,256,0.007654222349325816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,256,0.004990222139490975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,128,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,128,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,128,0.007666666474607255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,32,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,2048,0.011063999599880643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,64,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,65536,0.03725511166784499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,768,0.008656000097592672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,65536,0.051903999514049955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,256,0.0058017778727743365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,51200,0.03057866626315647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,65536,0.13311021857791477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,64,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,16384,0.014878221684032016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,51200,0.04455911119778951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,3072,0.007646222081449296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,51200,0.041026665104760066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,12288,0.013632888595263163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,16384,0.01755288905567593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,51200,0.10608088970184326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,16384,0.03940000136693319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,32,0.005000000198682149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,12288,0.01386577801571952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,65536,0.055092444022496544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,12288,0.01389955480893453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,10240,0.01242844429281023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,8192,0.012042666474978128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,10240,0.012301333248615265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,8192,0.011045333411958484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,8192,0.010770666930410596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,7168,0.011147555377748279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,16384,0.01665511065059238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,8192,0.023379556006855432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,6144,0.010413332945770687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,7168,0.01038933296998342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,7168,0.009639110830095079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,7168,0.0206631107462777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,5120,0.009688888986905416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,6144,0.009708444277445475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,6144,0.018914666440751817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,5120,0.008093333078755273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,5120,0.009016888837019602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,4096,0.008496888809733922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,4096,0.007376000285148621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,10240,0.01309866706530253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,3584,0.007629333270920648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,4096,0.008319999608728621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,12288,0.031207998593648274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,3584,0.007000889215204451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,3072,0.0070159998204973005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,5120,0.016541333662139047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,3072,0.006767999794748094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,4096,0.015021334091822306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,2560,0.005658666706747479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,6144,0.008995555341243744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,2560,0.006608888920810487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,3584,0.013790222505728403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,2048,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,2560,0.007357333269384172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,2048,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,3072,0.007669332954618666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,2048,0.006981333096822103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,1536,0.004618666652176115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,2048,0.011925333076053195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,3584,0.007976888782448238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,1536,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,1024,0.003972444269392225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,1536,0.010209778116808997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,1024,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,3072,0.013053333593739403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,1024,0.009369778136412302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,768,0.005607999861240387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,768,0.008571555217107138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,768,0.006259555617968242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,1536,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,512,0.0052764444715446895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,1024,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,768,0.0038897775941424896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,512,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,256,0.0059004442559348206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,256,0.005032000028424793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,256,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,512,0.0035413333939181436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,10240,0.027248000105222065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,64,0.00254133323000537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,128,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,32,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,128,0.007660444411966536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,65536,0.027275555663638647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,512,0.00795555528667238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,32,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,65536,0.04929600159327189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,51200,0.025069332785076563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,65536,0.13292622566223145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,2560,0.012144000165992312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,128,0.005747555444637935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,51200,0.0395671096112993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,51200,0.04427466789881388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,51200,0.10531644688712226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,256,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,12288,0.011873777541849347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,16384,0.016216889023780823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,16384,0.03861777649985419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,65536,0.05452444487147861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,10240,0.011056889262464313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,12288,0.01293688929743237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,12288,0.030293332205878362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,10240,0.011396444506115384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,8192,0.009669333696365356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,64,0.005031110925806893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,16384,0.016230222251680162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,8192,0.010022222167915767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,7168,0.009331555830107795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,8192,0.010839111275143094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,8192,0.023074666659037273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,7168,0.009389333426952362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,6144,0.008772444393899705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,7168,0.020974222156736586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,16384,0.012855110896958245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,10240,0.026399110754330952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,5120,0.008662222160233391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,6144,0.009706666900051964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,6144,0.01854222185081906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,5120,0.007979555262459649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,5120,0.009024000002278222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,5120,0.016531555189026725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,7168,0.01037599974208408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,10240,0.012247111234400006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,4096,0.008373333348168267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,3584,0.006959999601046245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,6144,0.008683555656009251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,3584,0.007234666910436418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,3584,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,3072,0.0063528890411059065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,128,0.0028684443483750024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,12288,0.01349866638580958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,2560,0.005629333357016246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,3072,0.006673777682913675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,3072,0.01314755529165268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,4096,0.014836443795098199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,2048,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,2560,0.012094222009181976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,2048,0.005975111077229182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,2048,0.011061333119869232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,1536,0.004641777939266629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,3072,0.007756444315115611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,4096,0.0074675555030504865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,1536,0.006129777679840724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,2560,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,1536,0.01036533299419615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,1024,0.005632888939645555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,3584,0.013807999591032663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,1024,0.009349333743254343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,1024,0.0063368889192740125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,2048,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,768,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,768,0.005587555468082428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,512,0.003519111209445529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,768,0.009015111459626092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,2560,0.006490666833188798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,512,0.005976888868543837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,256,0.0031902222997612427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,768,0.0035413333939181436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,256,0.004950222041871813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,256,0.005917333480384614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,128,0.002887111157178879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,1024,0.003926222109132343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,512,0.005280888742870755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,128,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,512,0.008395555946562026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,4096,0.007634667058785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,32,0.0028524444335036804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,64,0.004964444372389051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,768,65536,0.024752888414594863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,256,0.007670222057236566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,32,0.004937777916590373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,768,51200,0.02000266644689772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,65536,0.05436266793145073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,65536,0.13193421893649632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,64,0.0028391112056043413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,51200,0.03933066791958279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,51200,0.04348355531692505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,128,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,51200,0.1055484414100647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,128,0.005962666538026597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,16384,0.014820444915029736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,16384,0.03906044363975525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,12288,0.012454222473833295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,1536,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,12288,0.030188444587919448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,10240,0.011391110718250275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,768,16384,0.011904000408119626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,10240,0.026202667090627883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,10240,0.012137778103351593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,768,12288,0.01037333326207267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,16384,0.016170667277442086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,768,7168,0.008647110727098253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,8192,0.010013333625263637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,8192,0.02256711158487532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,65536,0.04882311158710056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,768,6144,0.008038221961922115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,7168,0.01036533299419615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,7168,0.02051555613676707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,6144,0.009635555247465769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,768,8192,0.009203555683294931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,6144,0.018560889694425795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,8192,0.010712888505723743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,5120,0.008029333419269985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,768,4096,0.007677333222495184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,5120,0.01621866722901662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,12288,0.013437333206335703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,7168,0.009407111340098912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,4096,0.008332444561852349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,4096,0.007369777394665613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,4096,0.015004444453451367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,6144,0.008835555778609382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,768,5120,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,3584,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,3584,0.014123555686738757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,3072,0.006702222343948152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,768,2560,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,3072,0.013108444710572561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,768,10240,0.010427555276287926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,768,3584,0.0070168889231152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,768,2048,0.0052284445199701525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,2560,0.012087110843923358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,768,3072,0.006314666734801398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,3584,0.006989333364698622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,2048,0.006261333409282897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,2048,0.0069982219073507525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,2048,0.011117333339320289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,3072,0.007688889073001013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,768,1024,0.0039048890272776284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,5120,0.008691555923885768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,1536,0.010422222316265106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,2560,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,1024,0.0063288890653186375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,1024,0.00904799997806549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,768,0.005585777676767773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,768,0.00835111074977451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,768,0.00600355573826366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,768,512,0.0032195556494924757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,1536,0.006735111276308696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,512,0.005315555466545953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,768,256,0.003206222214632564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,1024,0.005636444522274866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,256,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,768,1536,0.004573333180612988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,256,0.007694222033023834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,256,0.005664000080691443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,1536,0.00591822216908137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,128,0.004953777624501122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,768,64,0.0028417776856157514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,128,0.007727999654081132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,768,32,0.0028408887899584244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,768,768,0.003560888684458203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,512,65536,0.022241777843899194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,64,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,32,0.004965333475006951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,768,128,0.0028826666788922418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,65536,0.04785866538683573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,512,51200,0.018205333087179396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,65536,0.053842667076322764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,65536,0.1328213347329034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,512,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,51200,0.03870133227772183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,51200,0.10519822438557942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,16384,0.014861333701345654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,16384,0.01603822244538201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,512,12288,0.008770667016506195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,128,0.005704000178310607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,2560,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,12288,0.013479111095269522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,512,10240,0.009114666945404476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,16384,0.038020445240868464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,10240,0.011411555939250521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,512,8192,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,10240,0.012127111355463663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,10240,0.02607288956642151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,12288,0.012430222498046027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,512,7168,0.008023999631404877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,8192,0.009687999884287516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,8192,0.022346667117542688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,12288,0.0299911101659139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,512,6144,0.008480889101823172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,7168,0.009376889301670922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,7168,0.01033866653839747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,7168,0.020267556111017864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,512,5120,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,6144,0.009641778137948778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,6144,0.00869333330127928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,6144,0.018558222386572096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,512,4096,0.007346666521496243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,5120,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,512,0.007703111403518253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,5120,0.016749334004190233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,512,3584,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,4096,0.008349333372380998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,4096,0.014540443817774454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,3584,0.00702400008837382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,512,3072,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,3584,0.013762666947311826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,3072,0.006680888848172293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,5120,0.009009777671760982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,512,2560,0.005919999960396025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,3072,0.012902222573757172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,3072,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,8192,0.010848888920413123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,512,2048,0.005054222212897407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,4096,0.0075137780772315125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,2560,0.006350222147173352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,2560,0.01202933324707879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,512,1536,0.004228444563017951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,2048,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,2048,0.01111644423670239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,1536,0.00665866666369968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,1536,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,512,1024,0.003903999924659729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,1536,0.010031999813185798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,51200,0.0389013323518965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,512,768,0.0037271111375755737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,1024,0.006317333214812809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,1024,0.0053226666318045715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,2560,0.007322666545708974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,1024,0.009062222308582729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,768,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,768,0.008676444490750631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,512,0.005326222214433882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,512,256,0.0029111111329661477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,512,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,512,0.007725333174069722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,256,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,512,16384,0.01071466671095954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,256,0.004997333304749595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,256,0.007679111427730984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,512,512,0.0032479998966058097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,512,64,0.002522666628162066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,3584,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,512,32,0.0027502222607533136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,64,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,256,65536,0.01549600064754486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,32,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,512,128,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,65536,0.047260443369547524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,65536,0.047008888589011304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,128,0.005643555687533484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,768,0.005324444423119227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,2048,0.006239999913507038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,51200,0.036335110664367676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,51200,0.10439022382100423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,16384,0.014767110347747803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,16384,0.015855110353893705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,256,12288,0.009732444253232744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,256,51200,0.013451555536852943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,128,0.004971555537647671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,128,0.007853333320882585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,256,10240,0.009012444151772393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,256,16384,0.009413333402739631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,12288,0.02983199887805515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,10240,0.011389333340856763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,10240,0.02609155575434367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,256,8192,0.008300444318188561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,10240,0.012060444388124677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,65536,0.13083377149369982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,8192,0.009725333087974125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,256,7168,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,12288,0.013406222065289816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,8192,0.01072266697883606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,7168,0.010109333528412713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,7168,0.009352888498041365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,7168,0.020287111401557922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,16384,0.03797511259714762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,6144,0.0085235552655326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,256,5120,0.00702400008837382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,6144,0.009513777991135916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,8192,0.02232711182700263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,5120,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,256,4096,0.006607111129495833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,5120,0.016549333930015564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,256,6144,0.007335999773608313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,256,3584,0.006303999986913469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,4096,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,4096,0.014855111638704935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,3584,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,256,3072,0.006810666786299811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,3584,0.013786666923099093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,12288,0.012451555165979596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,51200,0.03804622093836466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,256,2560,0.0063324446479479475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,5120,0.009009777671760982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,3072,0.012810666528013019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,4096,0.008345777789751688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,2560,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,256,2048,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,6144,0.018575999471876357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,2048,0.00701155596309238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,2048,0.011085333095656501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,3072,0.006641777853171031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,3072,0.007592889169851939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,1536,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,256,1024,0.00360000009338061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,1536,0.010049777726332346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,2560,0.006648889018429651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,1024,0.0053297777970631914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,1024,0.0063368889192740125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,1024,0.009375111096435124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,256,1536,0.004231111043029361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,768,0.0053324442770746015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,768,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,768,0.008332444561852349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,1536,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,512,0.00592533333433999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,256,256,0.0028400001012616684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,2560,0.012101333174440594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,512,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,3584,0.008026666939258575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,256,768,0.0035377778112888336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,256,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,256,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,256,0.007798222204049428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,256,64,0.002516444358560774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,128,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,256,512,0.0032586666444937387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,128,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,128,65536,0.015434667468070984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,64,0.004938666605287128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,512,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,65536,0.04417777723736233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,2048,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,65536,0.046244445774290294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,256,128,0.002883555574549569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,65536,0.12964267200893825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,128,16384,0.008997333546479544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,51200,0.035649776458740234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,256,32,0.0028444443725877335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,16384,0.014815110299322339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,16384,0.015839111473825242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,128,12288,0.007978666987684038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,16384,0.0380257765452067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,32,0.005079999979999331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,128,10240,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,12288,0.013416888813177744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,12288,0.01239111108912362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,12288,0.029879109727011785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,10240,0.012095111111799875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,51200,0.0373911095990075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,128,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,51200,0.10317689180374146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,128,7168,0.006635555376609166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,8192,0.02241866621706221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,7168,0.00926577796538671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,128,51200,0.014495111174053617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,7168,0.020632000433074105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,128,6144,0.006592888798978593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,128,8192,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,10240,0.011402666568756104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,128,5120,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,6144,0.009691555466916826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,6144,0.018259555101394653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,8192,0.009703111317422655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,5120,0.00798044436507755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,128,4096,0.006324444380071428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,5120,0.016543111867374845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,7168,0.010109333528412713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,4096,0.007377777662542131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,128,3584,0.006343110981914733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,4096,0.00835111074977451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,10240,0.02607822252644433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,6144,0.00871022211180793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,128,3072,0.005964444329341252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,3584,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,3584,0.007132444116804335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,3584,0.013815999858909182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,5120,0.008665777742862701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,128,2560,0.006491555521885554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,3072,0.01275022245115704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,2560,0.007295110987292395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,128,2048,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,2560,0.011959111524952782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,4096,0.014545778433481852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,2048,0.00692622239391009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,128,1536,0.004572444491916233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,2048,0.010966222319338055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,3072,0.007635555333561367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,1536,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,128,1024,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,1536,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,2560,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,128,768,0.0035413333939181436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,1024,0.005626666463083691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,1024,0.009363555245929295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,2048,0.006083555519580841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,768,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,768,0.009024000002278222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,3072,0.006996444529957241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,512,0.004982222285535601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,128,256,0.0028986665937635633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,1536,0.010131555298964182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,512,0.00833422193924586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,256,0.005660444498062134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,128,128,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,256,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,8192,0.010833777487277985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,768,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,128,0.004981333182917701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,128,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,128,32,0.002499555548032125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,64,0.005244444641802046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,512,0.005994666781690385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,32,0.004896000027656555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,64,51200,0.011090666883521609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,256,0.004912000149488449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,64,16384,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,65536,0.04421155651410421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,64,12288,0.007689778175618913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,51200,0.03577511178122626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,128,64,0.0025315555847353404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,16384,0.014485332700941296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,128,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,12288,0.012429333395428128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,64,7168,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,10240,0.0110880004035102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,64,65536,0.012142221960756512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,8192,0.009718221922715506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,64,5120,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,7168,0.00941422250535753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,64,4096,0.005640000104904175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,6144,0.008719111482302347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,64,3584,0.006322666588756774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,128,512,0.0032026666320032547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,64,3072,0.006127999888526068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,4096,0.007650666766696506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,64,2560,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,64,8192,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,64,2048,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,3072,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,64,6144,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,2048,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,64,1024,0.0038977778620190094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,1536,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,64,768,0.003560888684458203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,5120,0.008022222254011366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,1024,0.005315555466545953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,3584,0.007087111473083496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,64,256,0.0029297777348094513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,512,0.0052177777720822235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,64,128,0.0028942223224374983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,256,0.005034666508436203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,64,64,0.002558222247494592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,128,0.0046302220887608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,64,32,0.0026559999419583213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,64,0.004971555537647671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,32,65536,0.012469333079126147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,32,0.004875555634498596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,32,51200,0.010766222245163388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,64,1536,0.004581333448489507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,32,16384,0.007978666987684038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,1024,0.006262222097979651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,51200,0.035432000954945884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,32,12288,0.0073724447025193115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,16384,0.01479822231663598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,32,10240,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,768,0.005292444593376583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,32,8192,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,10240,0.011111111276679568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,32,7168,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,8192,0.009723555710580613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,32,6144,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,7168,0.009384888741705153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,32,5120,0.005977777971161737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,64,512,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,6144,0.008353778057628209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,5120,0.008038221961922115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,32,3584,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,4096,0.007340444458855524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,2560,0.00665155549844106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,32,3072,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,32,2560,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,3584,0.007284444239404466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,32,2048,0.00600000015563435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,3072,0.007000889215204451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,32,1536,0.004264889078007804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,2560,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,32,1024,0.003842666745185852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,64,10240,0.007288888924651676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,32,4096,0.005623111294375525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,32,512,0.0032479998966058097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,1536,0.005984000033802456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,32,256,0.0028764444092909494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,768,0.005496888938877318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,32,128,0.002570666579736604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,256,0.005127999931573868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,32,64,0.00254488881263468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,128,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,32,32,0.00253866674999396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,12288,0.012099555797047086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,48,32,768,0.0036142224238978494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,32,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,1024,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,65536,0.04424622323777941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,512,0.0052675555149714155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,16384,0.36240800221761066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,51200,0.4918648931715224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,16384,0.13588621881273058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,64,0.005070222334729301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,12288,0.12325778272416855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,12288,0.1021066639158461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,51200,0.39612889289855957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,10240,0.10439911153581406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,10240,0.096596446302202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,16384,0.16607733567555746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,2048,0.006274666637182236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,7168,0.16116266780429417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,8192,0.0884595579571194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,8192,0.08882221910688613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,10240,0.22683376736111113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,6144,0.13722488615247938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,7168,0.06178577740987142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,5120,0.1172177791595459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,6144,0.05336977707015144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,4096,0.09349066681332058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,5120,0.05557511250178019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,12288,0.27186844083997935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,4096,0.04625066783693102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,4096,0.0388648874229855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,7168,0.07454578081766765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,3584,0.04281244344181485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,3584,0.04912799927923414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,3072,0.07238311237759061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,2560,0.06061689058939616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,3072,0.030132446024152968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,3072,0.04266577627923754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,2048,0.05011466807789273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,2560,0.02647200061215295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,2560,0.036024888356526695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,3584,0.08503999974992539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,1536,0.038689778910742864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,2048,0.03085244364208645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,1536,0.018548443913459778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,6144,0.06576533450020684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,1536,0.023927999867333308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,1024,0.01761777698993683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,51200,1.1244817309909396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,768,0.014249778456158109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,2048,0.022425777382320825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,512,0.014365333649847241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,1024,0.02816533380084568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,256,0.00923111124171151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,768,0.014820444915029736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,512,0.011845333708657159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,128,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,256,0.01037333326207267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,768,0.02126311096880171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,256,0.008887999587588841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,64,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,128,0.01074044406414032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,128,0.008112000094519721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,64,0.010064000056849586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,32,0.012430222498046027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,512,0.012237333589129977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,1024,0.015071999695565967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,8192,0.1840284400516086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,32,0.006575111299753189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,65536,1.0806400510999892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,51200,0.39491910404629177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,51200,0.8482817543877496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,16384,0.2757493389977349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,5120,0.057977775732676186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,65536,0.4076382319132487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,65536,0.5208009084065756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,12288,0.20672178268432617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,16384,0.10725866423712836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,10240,0.1716568867365519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,12288,0.08493600289026897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,8192,0.14063378175099692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,12288,0.09932533237669204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,10240,0.07634399996863471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,7168,0.12301245000627305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,8192,0.07161421908272637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,6144,0.10614133543438381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,7168,0.06085422303941515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,51200,0.3199662102593316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,5120,0.09011289146211411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,6144,0.05337599913279215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,4096,0.07341777616076998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,5120,0.04907377892070346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,10240,0.08419289191563924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,3584,0.06421599785486858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,8192,0.05719733238220215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,4096,0.03257600135273404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,16384,0.13259732723236084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,3072,0.05628889136844211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,3584,0.035169776942994856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,3584,0.04037244452370538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,6144,0.04392533169852363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,3072,0.02573866645495097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,3072,0.0354053344991472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,2048,0.03847200009557936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,2560,0.026566222310066223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,2560,0.02991022335158454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,2048,0.01926666663752662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,2048,0.025813332862324182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,1024,0.02214844524860382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,1536,0.01997066703107622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,4096,0.03755733370780945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,7168,0.05063644382688734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,768,0.016528889536857605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,1024,0.013386666774749756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,1024,0.014873777826627096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,512,0.011715555356608497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,768,0.012058667010731168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,2560,0.0469413333468967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,256,0.007968000239796108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,512,0.01038933296998342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,128,0.0046302220887608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,512,0.010392888552612728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,5120,0.04531466629770067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,256,0.007530666887760162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,128,0.009691555466916826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,32,0.004239999999602635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,128,0.007347555624114142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,64,0.008801778157552084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,32,0.010785777535703448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,65536,0.35807556576199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,65536,0.17535200383928087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,256,0.009674666656388177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,1536,0.030151999659008447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,51200,0.28052621417575413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,64,0.0042408886882993905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,65536,0.15126310454474554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,1536,0.015936889582210116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,12288,0.06160000297758314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,16384,0.04114933477507697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,51200,0.12313332822587754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,51200,0.1331555578443739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,12288,0.03309422069125705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,12288,0.03702044486999512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,8192,0.04311911265055338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,10240,0.03085422184732225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,10240,0.03227644496493869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,8192,0.02425866656833225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,8192,0.02789688772625393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,6144,0.032933334509531655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,7168,0.021998221675554912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,16384,0.0800186660554674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,7168,0.024028443627887305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,5120,0.02840177880393134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,6144,0.019759999381171334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,6144,0.022184888521830242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,5120,0.021180444293551978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,5120,0.01889511115021176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,4096,0.015487111277050443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,4096,0.016153777639071148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,7168,0.03783644570244683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,3072,0.018814222680197824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,3584,0.01752000053723653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,3072,0.013153778182135688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,3072,0.015779554843902588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,2560,0.016042666302786935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,4096,0.023560000790490046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,10240,0.051719999975628324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,2048,0.013341333303186627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,16384,0.04851288927925957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,2560,0.012098666694429187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,2560,0.013800000150998434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,768,0.012510221865442065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,2048,0.010800888968838586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,1024,0.008566222257084316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,1536,0.009561777942710454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,1536,0.010395555860466426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,768,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,1024,0.008748444418112436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,768,0.00739555557568868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,768,0.0069822221994400024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,512,0.005039111193683413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,3584,0.020791111720932853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,1536,0.01069155583779017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,512,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,2048,0.012358222570684222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,3584,0.016629333297411602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,256,0.006637333167923822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,256,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,64,0.0028817777832349143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,128,0.00683822234471639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,32,0.0032311110860771606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,64,0.00620977787507905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,32,0.006637333167923822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,256,0.003546666767862108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,65536,0.26787643962436253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,512,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,128,0.003216888962520493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,51200,0.2102728949652778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,65536,0.14828622341156006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,16384,0.07512799898783366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,1024,0.008314666648705801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,51200,0.11521777841779922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,128,0.005564444594913059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,10240,0.04520533482233683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,12288,0.02976977825164795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,12288,0.032448000378078885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,8192,0.03815644317203098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,10240,0.02676444417900509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,10240,0.02790488799413045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,7168,0.0329111119111379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,8192,0.024259555670950148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,8192,0.02444355520937178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,6144,0.028581334484948054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,7168,0.019997333486874897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,65536,0.1255306667751736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,7168,0.021347555849287245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,5120,0.0247013337082333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,16384,0.037273777855767146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,6144,0.01905333333545261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,4096,0.02128444446457757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,16384,0.042027556233935885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,3584,0.018564444449212816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,4096,0.014523555835088095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,4096,0.014283556077215405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,3072,0.01630311045381758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,3584,0.015680889288584392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,3072,0.01239377756913503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,3072,0.014175999495718213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,2560,0.013811555173661975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,51200,0.10272266467412312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,2048,0.01183733344078064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,6144,0.01794311073091295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,2560,0.011342222491900126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,1536,0.0099973330895106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,2560,0.012919999659061432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,2048,0.00998577806684706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,1024,0.008302222523424361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,2048,0.011360000405046674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,1536,0.0087333329849773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,768,0.006865777903132969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,1536,0.009700444837411245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,1024,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,512,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,768,0.0074942219588491656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,768,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,256,0.003867555409669876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,512,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,512,0.0063182223174307085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,128,0.003580444388919406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,256,0.006664888726340399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,5120,0.01923022170861562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,64,0.003206222214632564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,128,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,128,0.005346666607591841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,3584,0.015384000208642749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,64,0.006009777800904379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,1024,0.007324444750944774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,32,0.0063493334584765965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,65536,0.23343910111321342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,12288,0.0583404435051812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,51200,0.1846382220586141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,65536,0.14390844768948025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,16384,0.06031466854943169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,256,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,51200,0.10957066880332099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,51200,0.09514488776524861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,12288,0.048230220874150596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,32,0.0032035555276605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,10240,0.04135466615358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,12288,0.028392000330819026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,12288,0.030790223015679255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,8192,0.03020177616013421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,10240,0.026001777913835313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,10240,0.026897778113683064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,7168,0.030823998981051978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,8192,0.021010667085647583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,8192,0.02307555576165517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,6144,0.026733333865801494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,7168,0.019304000669055514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,7168,0.020008888509538438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,16384,0.040307554933759905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,6144,0.01757600075668759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,6144,0.017904000149832833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,4096,0.01961866683430142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,5120,0.017871111631393433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,4096,0.013960000541475085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,4096,0.013795555465751223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,3584,0.017129777206314933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,65536,0.12224800056881374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,5120,0.016557332542207506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,3584,0.014212444424629211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,5120,0.023048889305856492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,3072,0.011892444557613797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,3072,0.012825777961148156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,5120,0.01590222285853492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,2048,0.011620444556077322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,2048,0.009185777770148383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,1536,0.009477333062224919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,3072,0.015372445185979208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,3584,0.013688888814714221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,1024,0.007703111403518253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,16384,0.03624355461862352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,1536,0.008332444561852349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,768,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,1024,0.007636444436179266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,1024,0.0069013333155049225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,2560,0.010718222293588849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,768,0.007466666400432587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,2560,0.011400000088744693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,768,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,512,0.00700800038046307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,512,0.00591822216908137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,2048,0.01036000003417333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,128,0.003532444437344869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,256,0.005426666802830166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,64,0.0032355555643637977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,1536,0.00865955568022198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,32,0.003539555602603488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,64,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,512,0.00490399988161193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,32,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,65536,0.1750346687104967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,51200,0.13924445046318903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,65536,0.09376000033484565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,256,0.00666133314371109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,16384,0.047898666726218335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,128,0.006789333290523953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,128,0.00535111129283905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,16384,0.023181334137916565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,12288,0.03732088870472378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,16384,0.028413332170910303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,10240,0.030471109681659277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,12288,0.018626666731304593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,256,0.004217777815130022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,12288,0.02180622186925676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,8192,0.026458667384253606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,10240,0.016564443707466125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,7168,0.02264533274703556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,8192,0.016345777445369296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,7168,0.01350044459104538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,7168,0.014647111296653748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,6144,0.019919999771647982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,6144,0.01257333325015174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,5120,0.017128888103697035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,10240,0.019010666343900893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,8192,0.014727999766667684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,4096,0.01481599940194024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,2560,0.013167111410035027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,4096,0.01069866700304879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,4096,0.010427555276287926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,6144,0.01350044459104538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,3584,0.011257777611414591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,3072,0.011788444386588203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,65536,0.06960799958970812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,5120,0.012979555461141797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,2560,0.01035733355416192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,51200,0.0709493358929952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,2560,0.00887288898229599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,2048,0.009008888569143083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,3584,0.010777778095669217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,2048,0.008652444514963362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,5120,0.01202666676706738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,2048,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,3072,0.008982222113344405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,51200,0.05630489190419515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,1536,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,1024,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,2560,0.009654222263230218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,1024,0.006589333216349284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,1024,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,768,0.006457777900828256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,768,0.0057839999596277876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,1536,0.007711999946170383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,512,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,512,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,1536,0.007069333559936947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,3584,0.013088000317414602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,256,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,128,0.0029288888391521242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,128,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,64,0.0028568889117903183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,128,0.005057777795526716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,32,0.0032035555276605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,65536,0.15881244341532388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,32,0.00591555568906996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,768,0.004600000050332812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,256,0.005195555587609609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,51200,0.12563378281063503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,512,0.0035342222286595237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,51200,0.06454666455586751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,16384,0.044922666417227856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,51200,0.05429333448410034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,64,0.005621333503060871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,12288,0.03567733367284139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,16384,0.02199111051029629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,16384,0.026230222649044458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,12288,0.01812355551454756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,12288,0.020202666521072388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,65536,0.0669120020336575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,8192,0.024736000431908503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,10240,0.015985776980717976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,7168,0.021295999487241108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,8192,0.014185777968830533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,8192,0.015498666299713982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,6144,0.01868355605337355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,7168,0.013773333695199756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,256,0.0033599999215867785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,6144,0.012475555141766867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,10240,0.030751999881532457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,6144,0.012764444781674279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,4096,0.014444443914625378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,5120,0.011407111254003314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,65536,0.08395644691255356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,5120,0.01148622234662374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,10240,0.01773599949147966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,4096,0.010049777726332346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,4096,0.010311110979980892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,3072,0.011408888631396823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,3584,0.008782222039169734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,2560,0.009825777676370408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,3584,0.01074577785200543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,3072,0.010054222411579555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,2560,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,3072,0.010415111151006486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,5120,0.01609688831700219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,1536,0.007616889145639207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,2048,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,2048,0.007690666450394525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,1024,0.005576000031497743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,1536,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,1536,0.006724444528420766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,768,0.004513777792453766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,1024,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,1024,0.006335111127959357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,3072,0.00851022203763326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,768,0.0063271112740039825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,768,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,256,0.0032266666077905228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,512,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,512,0.005616888817813661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,128,0.002879111096262932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,256,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,256,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,3584,0.012647111382749347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,128,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,128,0.00498844434817632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,7168,0.013181333740552267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,64,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,32,0.005623111294375525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,2048,0.008828444613350762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,65536,0.13830578327178955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,2560,0.009185777770148383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,51200,0.0981466637717353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,65536,0.06446133057276408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,65536,0.07572444279988606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,51200,0.052051554123560585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,51200,0.0583475563261244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,64,0.0028453332682450614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,16384,0.024130667249361675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,32,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,10240,0.026735110415352717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,12288,0.017476444443066914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,8192,0.02239911092652215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,512,0.0038817777401871155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,10240,0.01611022154490153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,8192,0.013737777868906656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,7168,0.020480889413091872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,16384,0.03719288773006863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,12288,0.03107288810941908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,7168,0.012936000194814471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,6144,0.01847822301917606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,7168,0.013017777767446307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,5120,0.01650399963061015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,6144,0.01184800018866857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,12288,0.018271999226676095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,6144,0.01198755535814497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,4096,0.012916444076432122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,5120,0.01072266697883606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,5120,0.01073066641887029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,4096,0.009347555538018545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,8192,0.014505777094099255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,3072,0.010839111275143094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,3584,0.010199999643696679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,16384,0.020956445071432326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,3072,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,3072,0.009513777991135916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,2048,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,2560,0.007670222057236566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,2560,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,1536,0.006907555378145642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,2048,0.007327111230956183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,4096,0.009854222337404886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,1536,0.006959999601046245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,1024,0.004600888739029567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,1536,0.006595555692911148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,768,0.004242666479614046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,1024,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,1024,0.005964444329341252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,512,0.0035626664757728577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,768,0.005618666609128316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,10240,0.015487111277050443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,256,0.0035386664999855887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,512,0.006117333140638139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,512,0.005429333282841577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,3584,0.011760888828171624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,128,0.0031848889258172777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,256,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,64,0.0029120000286234748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,128,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,32,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,128,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,65536,0.11586933665805393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,64,0.0052800000541739995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,768,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,51200,0.08572089009814793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,65536,0.06234755780961779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,65536,0.06969600253634982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,2048,0.007349333001507654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,51200,0.050308446089426674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,256,0.005221333354711533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,3584,0.008456000023418004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,12288,0.02718844513098399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,16384,0.020283554991086323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,10240,0.022846221923828125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,12288,0.01685333251953125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,2560,0.009400000174840292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,12288,0.016692444682121277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,8192,0.019276445110638935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,10240,0.0151555554734336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,10240,0.014921777778201632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,8192,0.01332800007528729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,8192,0.013791110780504016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,6144,0.015809777710172866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,7168,0.0124177775449223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,51200,0.05283111002710131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,5120,0.015643555257055495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,6144,0.011061333119869232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,16384,0.021497777766651575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,6144,0.011515555282433828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,4096,0.012431999875439538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,5120,0.00997511131895913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,32,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,5120,0.010384889112578498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,4096,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,3072,0.010240000155236986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,3584,0.00831733312871721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,16384,0.0336408880021837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,2560,0.008963555925422246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,3072,0.007978666987684038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,7168,0.012076444096035428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,2048,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,2560,0.007635555333561367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,2560,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,1536,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,2048,0.007296000089910295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,2048,0.0069573331210348345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,1024,0.004281777888536453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,3584,0.011071999867757162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,4096,0.009398221969604492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,1536,0.006311999840868845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,768,0.00424888895617591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,1024,0.006305777778228124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,3584,0.009694221946928237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,768,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,3072,0.008727110922336578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,512,0.006010666489601135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,512,0.005299555758635203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,256,0.0035217776894569397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,256,0.005972444597217772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,1536,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,256,0.004972444640265571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,7168,0.01754577789041731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,64,0.0028773333049482773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,128,0.005986666513813867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,512,0.003543111185232798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,128,0.004942222187916438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,768,0.005625777774386936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,64,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,32,0.005305777821275923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,51200,0.07113155391481188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,65536,0.06287378072738647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,128,0.003228444399105178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,16384,0.028314666615592107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,51200,0.048455112510257296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,12288,0.022720000810093347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,1024,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,16384,0.018948444061809115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,10240,0.02010933392577701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,65536,0.08761066860622829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,12288,0.015490666031837463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,10240,0.014488889111412896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,8192,0.01759999990463257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,10240,0.014114666316244336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,7168,0.015535111228624979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,8192,0.01293333371480306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,51200,0.04796266555786133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,6144,0.012647999657524956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,7168,0.011752000285519494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,16384,0.019697777099079557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,5120,0.011129777464601727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,6144,0.010703110860453712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,6144,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,4096,0.010045333041085137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,5120,0.009062222308582729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,5120,0.009668444593747457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,3584,0.009015111459626092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,4096,0.008345777789751688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,4096,0.00869688888390859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,3072,0.008184889124499427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,3584,0.008254222571849823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,7168,0.011416888899273343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,32,0.002891555635465516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,3072,0.007731555236710443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,12288,0.01632355484697554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,2560,0.007332444190979004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,2560,0.006948444578382704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,65536,0.05968799855973986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,2048,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,8192,0.012829333543777466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,3584,0.008045333127180735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,1024,0.0042088888585567474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,1536,0.006653333289755716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,1536,0.005964444329341252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,768,0.0038604442444112566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,2560,0.007378666765160031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,1024,0.005583999885453119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,512,0.0035662220584021676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,2048,0.005988444305128521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,768,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,256,0.0031991110493739447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,512,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,512,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,1536,0.005015999906592899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,256,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,256,0.004933333231343163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,64,0.0028657778683635923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,128,0.00590844452381134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,32,0.002862222285734283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,1024,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,3072,0.007676444119877285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,65536,0.07566133472654554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,768,0.0063271112740039825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,51200,0.06077688932418823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,65536,0.05990311172273424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,2048,0.00703288863102595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,128,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,16384,0.025966222087542217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,51200,0.047228445609410606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,51200,0.04685422115855747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,12288,0.020628444022602506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,16384,0.019344889455371432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,10240,0.01787644459141625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,12288,0.01610577768749661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,12288,0.015065777632925244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,64,0.005059555586841372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,8192,0.01605866683853997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,10240,0.01366222235891554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,7168,0.014624888698259989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,8192,0.012437333663304647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,8192,0.0122097780307134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,6144,0.014069333672523499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,7168,0.01108888867828581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,6144,0.009739555418491364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,6144,0.01036711119943195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,5120,0.010713777608341642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,10240,0.014463999205165438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,32,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,5120,0.009213333328564962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,5120,0.009069333473841349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,7168,0.011136000355084738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,4096,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,128,0.004949333353175057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,3584,0.007654222349325816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,3072,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,3584,0.008203555312421586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,3072,0.007879111501905654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,65536,0.05882844660017225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,3072,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,4096,0.00832177781396442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,2560,0.007335111498832703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,3584,0.008738666772842407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,2560,0.006664888726340399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,1536,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,2048,0.0070053330726093715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,16384,0.018836443622907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,2048,0.006275555739800136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,1536,0.006673777682913675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,768,0.003903999924659729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,1024,0.005653333332803514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,4096,0.009714666340086196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,2560,0.007254222200976477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,768,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,512,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,512,0.0053075556125905775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,256,0.003216888962520493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,256,0.0058311112225055695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,128,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,1536,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,128,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,512,0.0035502223504914176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,768,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,32,0.0028897778441508612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,2048,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,64,0.004948444250557157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,65536,0.06685066885418363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,32,0.0053182223604785065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,1024,0.0041991112132867174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,65536,0.05754488706588745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,65536,0.057160000006357826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,1024,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,16384,0.021915555000305176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,51200,0.046205331881841026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,12288,0.018953777021831937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,16384,0.01795111099878947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,12288,0.015837333268589444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,10240,0.01585155559910668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,51200,0.05379022161165873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,256,0.004966222163703707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,10240,0.013129777378506131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,8192,0.014128888646761576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,51200,0.04494666722085741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,8192,0.011744889120260874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,7168,0.013312888642152151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,8192,0.01202044470442666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,6144,0.011844444606039258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,7168,0.010746666954623329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,7168,0.010915555887752108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,12288,0.014470222923490735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,6144,0.010051555931568146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,5120,0.008969777988062965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,10240,0.014119111829333834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,128,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,16384,0.01925511161486308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,3584,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,4096,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,3584,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,3584,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,3072,0.0076657773719893555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,3072,0.007682667010360294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,5120,0.008962666822804345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,3072,0.007170666423108842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,64,0.002879111096262932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,2048,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,4096,0.007696888513035244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,2048,0.0070293330483966405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,5120,0.011162666810883416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,1536,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,2560,0.0069422221018208405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,1024,0.004282666577233208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,1536,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,2560,0.007641777396202087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,768,0.0038844446341196695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,1024,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,1024,0.005588444570700328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,512,0.003539555602603488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,768,0.006230222268237009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,6144,0.010071111222108206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,256,0.0031822222388452957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,512,0.005247999810510212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,4096,0.009200889203283522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,128,0.0028684443483750024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,2560,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,128,0.004983110974232356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,128,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,64,0.002834666727317704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,1536,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,32,0.0028524444335036804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,64,0.0052355556852287715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,65536,0.057479110029008657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,32,0.0053182223604785065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,65536,0.056800888644324414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,51200,0.04881599876615736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,2048,0.0063075555695427795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,512,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,16384,0.019730667273203533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,51200,0.043395555681652494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,51200,0.045548445648617215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,12288,0.01718933383623759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,256,0.005919999960396025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,16384,0.018943111101786297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,16384,0.017435555656750996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,12288,0.015146666102939181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,65536,0.05538400014241537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,10240,0.013119999733236102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,10240,0.01295111080010732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,768,0.005428444594144821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,8192,0.011405333048767514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,7168,0.012420444852775998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,8192,0.011532444092962476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,6144,0.01128355579243766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,7168,0.010616000327799054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,256,0.005120888766315249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,5120,0.010682666467295753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,6144,0.009516444471147325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,4096,0.009034666750166152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,5120,0.00905955582857132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,5120,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,8192,0.013154666456911298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,4096,0.008447111480765873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,4096,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,3072,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,3584,0.008112000094519721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,3584,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,2560,0.0069013333155049225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,3072,0.007781333393520779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,7168,0.01072622256146537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,3072,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,6144,0.009702222214804756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,2560,0.0064604443808396654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,1536,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,2048,0.006266666783226862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,12288,0.014167111780908374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,3584,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,1536,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,1536,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,768,0.0035537779331207275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,1024,0.006299555715587404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,1024,0.005617777920431561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,512,0.0032257777121331957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,768,0.005504888792832692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,2048,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,10240,0.014925332532988654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,256,0.00319377767542998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,512,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,512,0.005280888742870755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,256,0.005728888842794631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,256,0.005239999956554837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,1024,0.0042257776690853965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,128,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,32,0.0028195555011431375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,64,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,768,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,65536,0.050831112596723765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,32,0.00500266667869356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,2560,0.007666666474607255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,51200,0.04016800059212579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,65536,0.05560266971588135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,128,0.0028773333049482773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,16384,0.017200888858901132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,51200,0.0448017782635159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,51200,0.04170755545298258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,128,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,16384,0.016908443636364408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,10240,0.013792888985739814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,12288,0.013931555880440606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,10240,0.012450666891203986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,10240,0.012541333006487953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,8192,0.013129777378506131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,2048,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,8192,0.011109333071443768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,7168,0.011554666691356234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,12288,0.015240889456537036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,16384,0.018509333332379658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,7168,0.010388444695207808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,7168,0.009830222361617619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,5120,0.010057777994208867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,6144,0.009015111459626092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,5120,0.009011555876996782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,4096,0.009007111191749573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,65536,0.05269333389070299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,8192,0.011048888994587792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,3584,0.007744889292452071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,4096,0.008455110920800103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,64,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,6144,0.01071111112833023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,3072,0.006903999795516332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,12288,0.01463644372092353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,3584,0.008061333662933773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,6144,0.009869332942697737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,2560,0.005991111199061076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,3072,0.00792711145348019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,2048,0.0052888890107472735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,3072,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,5120,0.008090666598743862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,2560,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,1536,0.004598222259018156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,2048,0.006983111302057902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,1024,0.004232888834344016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,1536,0.006706666615274217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,1536,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,4096,0.007329777710967594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,768,0.0038942222793896994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,1024,0.005626666463083691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,1024,0.006343110981914733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,512,0.003552888830502828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,768,0.006170666466156642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,256,0.0031911111954185697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,512,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,512,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,2560,0.007628444168302748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,128,0.0028942223224374983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,256,0.005623999983072281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,64,0.002537777854336633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,128,0.005973333285914526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,2048,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,32,0.0028453332682450614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,128,0.004960000101062986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,65536,0.038635555240843035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,64,0.004951111144489712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,51200,0.03112799922625224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,65536,0.055161777469846934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,65536,0.05054310957590739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,3584,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,16384,0.015142222245534262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,768,0.005294222384691238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,12288,0.013519110778967539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,16384,0.017595556047227647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,16384,0.016569778323173523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,256,0.0052133335007561584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,10240,0.01218844453493754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,12288,0.013455111119482251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,8192,0.010693333215183683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,10240,0.012429333395428128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,7168,0.01072533345884747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,10240,0.012096888489193387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,32,0.00499288903342353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,6144,0.010812444819344414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,7168,0.01037333326207267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,7168,0.009529777699046666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,51200,0.04450577828619215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,5120,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,51200,0.04012177719010247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,6144,0.008752000000741746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,4096,0.008354666332403818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,5120,0.00902133352226681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,3584,0.0074577778577804565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,12288,0.013780444032616086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,4096,0.007418666448858049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,3072,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,3584,0.008041777544551426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,8192,0.01074577785200543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,2560,0.00591466658645206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,3072,0.007817777494589487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,3072,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,2048,0.005215999980767568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,6144,0.009724444813198512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,2048,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,2048,0.006108444597986009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,4096,0.008340444829728868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,1536,0.006596444381607904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,1024,0.003923555629120933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,8192,0.01035288886891471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,1024,0.006274666637182236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,2560,0.007461333440409766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,2560,0.0064497776329517365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,768,0.0061048890153567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,512,0.003524444583389494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,1536,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,512,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,256,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,1024,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,256,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,128,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,1536,0.004591111093759537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,128,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,64,0.0026400000270869997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,3584,0.007300444775157505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,32,0.002513777878549364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,64,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,768,0.0038897775941424896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,65536,0.030239999294281006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,32,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,51200,0.02444355520937178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,256,0.00509333320789867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,65536,0.054640889167785645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,65536,0.048370665974087186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,16384,0.01276622215906779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,51200,0.04411377840571933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,128,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,512,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,12288,0.010620444185203977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,12288,0.013343999783198038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,12288,0.012402666939629449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,5120,0.008162666526105668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,10240,0.011342222491900126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,8192,0.009342222577995723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,10240,0.012117333710193634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,51200,0.03906755646069845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,7168,0.00887111077706019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,16384,0.016175111134847004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,7168,0.010371555884679159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,6144,0.007976888782448238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,7168,0.009351111120647853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,6144,0.009678222239017487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,768,0.005478222337034013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,6144,0.00870577742656072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,5120,0.009000889129108852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,4096,0.007626666790909237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,8192,0.01109333336353302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,16384,0.015622221761279635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,3584,0.007076444725195567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,4096,0.008365333080291748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,3072,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,10240,0.01071111112833023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,3584,0.007191110816266801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,3072,0.007607111500369177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,2560,0.005633777628342311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,5120,0.007968000239796108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,2048,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,2560,0.007279111279381647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,2560,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,1536,0.004651555584536659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,2048,0.0069448889957533936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,2048,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,8192,0.0099653336736891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,1024,0.0038782221575578055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,1536,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,768,0.0037324445115195382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,1536,0.005973333285914526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,1024,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,3584,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,512,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,768,0.005909333212508096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,256,0.0028986665937635633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,768,0.0052764444715446895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,3072,0.006844444407357111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,512,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,128,0.0029022221763928733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,256,0.005590222362014983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,64,0.002568888788421949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,256,0.004940444396601783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,5120,0.00817599975400501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,128,0.004905777672926585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,768,65536,0.025075554847717285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,64,0.004893333133724001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,32,0.004975999808973736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,4096,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,768,51200,0.020907556017239887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,65536,0.05443111062049866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,65536,0.04782844583193461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,768,16384,0.011050666371981302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,512,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,768,12288,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,16384,0.016010666886965435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,16384,0.01495822270711263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,768,10240,0.012083555261294046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,32,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,12288,0.012418666647540199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,768,8192,0.010713777608341642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,10240,0.012070222033394707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,10240,0.011135111252466837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,768,7168,0.009683555199040307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,1024,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,7168,0.010039110978444418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,51200,0.0384133325682746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,128,0.005889777921968036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,6144,0.00867377801073922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,768,5120,0.008167111211352879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,8192,0.010689777632554373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,8192,0.00980177770058314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,5120,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,51200,0.04376533296373156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,768,6144,0.009007111191749573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,768,3584,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,12288,0.013467555244763693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,6144,0.009731555150614845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,768,3072,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,3072,0.007010666860474481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,7168,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,768,2560,0.005583999885453119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,3584,0.007237333390447829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,2560,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,768,2048,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,768,4096,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,4096,0.008392000363932716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,2048,0.0069884442620807225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,2048,0.006224000205596288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,3584,0.007714666426181793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,1536,0.006653333289755716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,768,1024,0.0038897775941424896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,5120,0.00869333330127928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,4096,0.007355555891990662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,768,768,0.003540444291300244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,1024,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,768,1536,0.004557333472702238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,2560,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,768,0.0053182223604785065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,1536,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,512,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,3072,0.007626666790909237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,256,0.00499288903342353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,768,128,0.0028826666788922418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,768,0.005917333480384614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,128,0.005648889061477449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,768,64,0.0026355555488003623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,768,32,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,512,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,64,0.0048942222363419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,512,65536,0.017723555366198223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,32,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,512,51200,0.014937778313954672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,1024,0.005625777774386936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,65536,0.0471937788857354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,256,0.005869333528810077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,51200,0.038768000072903104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,51200,0.03779822256830003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,512,12288,0.010447111394670276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,16384,0.015903110305468243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,128,0.004912000149488449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,512,10240,0.009512888888518015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,12288,0.012115555504957834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,768,256,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,512,8192,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,10240,0.012115555504957834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,10240,0.011172444456153445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,65536,0.05242222216394213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,512,7168,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,8192,0.01072266697883606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,8192,0.009917333722114563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,7168,0.010330666270520952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,7168,0.009365333451165093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,768,512,0.0032640000184377036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,6144,0.008681777450773451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,12288,0.013418667018413544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,512,4096,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,5120,0.008202667037645975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,4096,0.008365333080291748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,512,3584,0.0069413334131240845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,512,16384,0.01073066641887029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,3584,0.00810222244924969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,512,5120,0.007985778152942657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,512,3072,0.006269333263238271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,16384,0.014853333433469137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,512,2560,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,3072,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,5120,0.009020444419648912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,512,2048,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,2560,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,2560,0.006653333289755716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,2048,0.0069520001610120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,4096,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,2048,0.006256000035338932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,512,6144,0.007325333025720384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,512,1024,0.0038737778862317405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,1536,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,512,768,0.003556444413132138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,3584,0.007059555914666917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,1024,0.006530666516886816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,512,512,0.0032195556494924757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,1024,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,768,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,512,256,0.0029031110720502008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,512,0.005889777921968036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,3072,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,512,128,0.0028764444092909494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,256,0.005018666800525454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,512,1536,0.004613333278232151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,512,64,0.00258577780591117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,128,0.005819555371999741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,128,0.00490399988161193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,1536,0.00592533333433999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,6144,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,64,0.004958222309748332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,32,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,768,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,65536,0.04690488841798571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,65536,0.0441599984963735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,512,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,256,16384,0.009943111075295342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,256,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,256,12288,0.009712888962692684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,16384,0.01588000026014116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,16384,0.01459377838505639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,512,32,0.0028560000161329904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,256,10240,0.008990222381220924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,12288,0.012101333174440594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,256,65536,0.01512888901763492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,256,8192,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,10240,0.01204622205760744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,10240,0.011428444749779172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,256,51200,0.013425778183672162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,256,7168,0.009354666703277165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,8192,0.010025777750545079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,51200,0.03768444392416213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,51200,0.035537779331207275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,7168,0.009379555781682333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,256,5120,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,6144,0.009535111486911774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,12288,0.013387555877367655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,256,4096,0.007341333561473423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,5120,0.008916444248623317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,5120,0.007991111112965478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,256,3584,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,4096,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,4096,0.00740444411834081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,256,3072,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,3584,0.007727111379305522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,3584,0.007017778025733099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,7168,0.01016533292002148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,256,2560,0.006270222365856171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,3072,0.006999111009968652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,256,2048,0.005241777747869492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,6144,0.008681777450773451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,2048,0.0070017774899800616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,2048,0.006229333165619109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,8192,0.010715555813577441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,256,1024,0.0039022221333450745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,256,6144,0.00832444429397583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,1024,0.006396444307433234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,1024,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,256,768,0.003544888976547453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,768,0.006107555495368109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,256,1536,0.004567111117972268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,512,0.005320888840489917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,1536,0.00664533343580034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,3072,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,256,256,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,512,0.005985777825117111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,256,128,0.002924444360865487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,2560,0.007325333025720384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,256,0.005619555711746216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,2560,0.006671111202902264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,256,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,128,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,256,32,0.002842666581273079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,128,0.004976888911591636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,256,512,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,64,0.004935111022657818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,32,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,65536,0.04622488882806566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,65536,0.043507556120554604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,1536,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,128,16384,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,51200,0.037610666619406805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,128,12288,0.00831200016869439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,16384,0.015867556134859722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,16384,0.014607111612955729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,256,64,0.002883555574549569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,128,10240,0.008656889200210571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,12288,0.01218488895230823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,128,8192,0.00814577771557702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,768,0.005576888720194499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,10240,0.010851555400424533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,128,7168,0.007646222081449296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,8192,0.009848888549539778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,7168,0.010259555445777046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,128,6144,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,12288,0.01313777764638265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,6144,0.008660444782839881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,128,5120,0.006945778098371293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,128,51200,0.01421511173248291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,5120,0.00906488878859414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,128,4096,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,7168,0.009089777866999308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,4096,0.008019555773999956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,128,3584,0.006312888943486744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,128,65536,0.015096000499195524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,3584,0.007824000385072496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,128,3072,0.005595555735958948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,5120,0.008100444244013892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,51200,0.03530577818552653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,3072,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,4096,0.007367110914654202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,2560,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,10240,0.012097777591811286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,128,2048,0.005716444303592046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,3584,0.006991110742092133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,2048,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,2048,0.006275555739800136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,1536,0.006607111129495833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,128,1024,0.0038746665749284956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,1024,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,2560,0.007367110914654202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,1024,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,128,768,0.003547555456558863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,768,0.0053013331360287136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,1536,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,128,512,0.003216000066863166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,768,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,512,0.005961777849329843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,128,256,0.0029351111087534162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,8192,0.010896888871987661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,128,1536,0.00453244439429707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,128,128,0.00254844439526399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,256,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,128,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,3072,0.007640000018808577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,128,0.004943111290534337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,128,32,0.002571555475393931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,64,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,512,0.00509066672788726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,32,0.004592000196377437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,64,51200,0.01351288871632682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,256,0.004984000076850255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,65536,0.043172445562150746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,64,16384,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,64,12288,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,128,2560,0.005969777703285217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,16384,0.014519999424616495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,64,8192,0.006685333533419504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,6144,0.009725333087974125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,64,65536,0.016167999969588388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,8192,0.009709333380063375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,64,6144,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,7168,0.009368889033794403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,64,5120,0.006063111126422882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,51200,0.03495555453830295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,6144,0.00866844422287411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,5120,0.008000888758235509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,12288,0.012114666402339935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,4096,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,64,3072,0.005722666780153911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,3584,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,64,2560,0.006320888797442119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,3072,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,64,2048,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,2560,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,64,1536,0.004275555411974589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,2048,0.005995555470387141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,64,3584,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,10240,0.011044444309340583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,64,768,0.003528888854715559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,1024,0.005579555614127054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,64,512,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,768,0.005588444570700328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,64,256,0.002842666581273079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,512,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,64,128,0.0026168889469570587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,256,0.0053004444473319584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,64,1024,0.0038693332009845306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,128,0.00490488898422983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,64,7168,0.006612444503439798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,64,32,0.0028106667515304354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,32,65536,0.016174222032229107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,64,0.004951999833186467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,32,51200,0.013448889056841532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,32,0.004939555707905027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,32,16384,0.008043555749787224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,65536,0.04349066813786825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,128,64,0.0025840000145965153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,51200,0.034853332572513156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,32,10240,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,16384,0.01479555500878228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,64,64,0.002499555548032125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,32,8192,0.0069919998447100324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,12288,0.012103999654452005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,32,7168,0.006692444284756978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,1536,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,8192,0.00980533328321245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,7168,0.009358222285906473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,6144,0.00852177788813909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,32,4096,0.005642666584915585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,64,10240,0.007033777733643849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,32,3584,0.006117333140638139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,4096,0.007326222128338284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,32,3072,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,3584,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,32,6144,0.006342222293217977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,32,2560,0.005979555762476391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,32,2048,0.00565422202150027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,3072,0.006879999819729064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,32,1536,0.004576000074545543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,2560,0.006683555742104848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,32,1024,0.003912000192536248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,2048,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,32,768,0.0034391110142072043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,1536,0.005690666536490123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,32,512,0.0031964443624019623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,1024,0.005653333332803514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,32,256,0.002887111157178879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,768,0.005312888986534542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,32,128,0.002883555574549569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,10240,0.010834666589895884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,512,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,32,64,0.0025333333760499954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,32,5120,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,32,32,0.0028080000645584534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,64,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,64,4096,0.005576000031497743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,32,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,256,0.005299555758635203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,51200,0.4940142101711697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,128,0.005098666581842634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,51200,1.1169982486300998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,33,32,12288,0.007810666329330868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,16384,0.16568177276187473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,12288,0.2733528878953722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,51200,0.4285857677459717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,5120,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,12288,0.10512977838516235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,16384,0.15209333101908365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,10240,0.2310240003797743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,10240,0.1045759982532925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,12288,0.12322310606638591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,8192,0.18442932764689127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,12288,0.21729600429534912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,8192,0.0886497762468126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,8192,0.0729057788848877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,10240,0.18266933494144014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,16384,0.2764684359232585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,7168,0.07462310791015625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,16384,0.36484622955322266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,8192,0.1432151132159763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,6144,0.1386302179760403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,7168,0.06436533398098417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,5120,0.11761689186096191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,7168,0.12956178188323975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,6144,0.06535555256737603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,7168,0.1606702274746365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,6144,0.06578310992982653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,5120,0.052899556027518384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,6144,0.10862488879097833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,10240,0.0888213316599528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,5120,0.09421600235833062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,51200,0.8833351135253906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,4096,0.07684444718890719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,3584,0.046105778879589505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,3584,0.03699644406636556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,5120,0.05555377735031975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,4096,0.09483555290434097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,3584,0.07646666632758246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,4096,0.04629155662324694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,2560,0.05955288807551066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,3072,0.031585776143603854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,2560,0.03427911135885451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,2560,0.02755022214518653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,2048,0.047714667187796704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,2048,0.029542220963372126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,3072,0.041936000188191734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,2048,0.04661688870853848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,1536,0.03762133253945245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,3584,0.08407200045055813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,1536,0.019596444235907663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,1536,0.02302933401531643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,1024,0.026407111022207472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,1536,0.03875644339455499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,3072,0.07134044170379639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,768,0.020607999629444547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,1024,0.0170168893204795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,1024,0.030927111705144245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,768,0.01439377831088172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,4096,0.04063377777735392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,2048,0.023390221926901076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,3072,0.06515733400980632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,768,0.026733333865801494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,256,0.008355555435021719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,512,0.012120889292822944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,512,0.011797332929240333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,512,0.022600889205932617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,2560,0.06510044468773736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,256,0.009164444274372524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,256,0.020688888099458482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,128,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,64,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,128,0.01588888963063558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,768,0.014577777849303352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,64,0.010098666780524785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,32,0.012454222473833295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,128,0.0070053330726093715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,1024,0.017323555217848886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,128,0.010737777584128909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,512,0.01421333352724711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,32,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,65536,0.5216551356845432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,65536,1.1227084265814888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,256,0.010728889041476779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,51200,0.3942417833540175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,51200,0.8591929011874728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,16384,0.13256977664099798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,51200,0.35484978887769913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,12288,0.21192977163526747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,16384,0.11100622018178304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,16384,0.284076452255249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,12288,0.08442133002811009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,16384,0.23447465896606445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,65536,0.45053688685099286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,12288,0.0992399983935886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,12288,0.17201599809858534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,8192,0.1430248949262831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,10240,0.17578932974073622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,10240,0.14640977647569445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,8192,0.07201955715815227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,8192,0.05987289216783312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,10240,0.0845022201538086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,8192,0.11860799789428711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,7168,0.06074755721622043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,6144,0.107588443491194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,7168,0.05232266585032145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,7168,0.1056435571776496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,7168,0.12598666879865858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,6144,0.05327999922964308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,5120,0.09199021922217475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,6144,0.045759999089770846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,65536,0.9455493291219076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,6144,0.09301777680714925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,4096,0.07523733377456665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,5120,0.04155288802252875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,5120,0.04518577787611219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,3584,0.0649991101688809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,5120,0.07840177747938368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,4096,0.06664178106519911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,3584,0.03872266742918227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,3584,0.030560887522167627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,3584,0.05999822086758084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,10240,0.07259644402398004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,3072,0.02646488944689433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,2560,0.04810488886303372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,51200,0.7256399790445963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,4096,0.0373911095990075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,2048,0.03925688730345832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,2560,0.028832889265484277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,2560,0.02344977855682373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,3072,0.056998219754960805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,2048,0.01997066703107622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,1536,0.03055466545952691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,3072,0.034040000703599714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,1536,0.01941422290272183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,1536,0.016757332616382174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,1024,0.021983999345037673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,4096,0.034090667963027954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,1024,0.01367200000418557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,768,0.016857778032620747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,1024,0.027064000566800434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,2560,0.05299199952019585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,2048,0.040404445595211454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,512,0.011726222104496427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,768,0.013231111069520315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,3072,0.05323911044332716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,1536,0.03400444322162204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,256,0.007329777710967594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,512,0.01982311076588101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,1024,0.01463733282354143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,256,0.007684444387753804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,128,0.00423200014564726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,256,0.018106665876176622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,768,0.01276711126168569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,128,0.007465778125656976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,768,0.023220444718996685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,128,0.009675555759006077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,32,0.003882666842805015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,512,0.010508444574144153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,64,0.00903733323017756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,32,0.010775999890433418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,65536,0.3579866621229384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,256,0.009657777845859528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,2048,0.02716355522473653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,65536,0.17921245098114014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,51200,0.2673128975762261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,64,0.003539555602603488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,128,0.014352889524565803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,51200,0.13298577732510036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,512,0.010427555276287926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,16384,0.042862223254309766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,12288,0.061299555831485324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,16384,0.05501155720816719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,12288,0.03689422210057577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,51200,0.13709867000579834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,10240,0.05158133308092753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,51200,0.12184088759952122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,10240,0.03197777933544583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,10240,0.029299557209014893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,10240,0.0358240008354187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,8192,0.044030222627851695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,65536,0.15529155731201172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,8192,0.02878311276435852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,7168,0.037736889388826154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,16384,0.04796977837880453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,8192,0.031799111101362444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,7168,0.022648000054889258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,6144,0.03259377678235372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,7168,0.027858667903476294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,65536,0.17967466513315836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,6144,0.021669333179791767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,6144,0.024992000725534227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,6144,0.0207360006040997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,5120,0.028451555305057104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,5120,0.018783110711309645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,5120,0.0222746663623386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,4096,0.024432889289326135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,5120,0.019367999500698514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,16384,0.08071821928024292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,4096,0.016191111670600045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,3584,0.020797333783573575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,4096,0.016179554992251925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,4096,0.019356444478034973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,3584,0.01497333414024777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,3584,0.01722311145729489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,3072,0.0185253338681327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,3584,0.01806311143769158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,3072,0.013468444347381592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,2560,0.015779554843902588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,3072,0.01664266652531094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,12288,0.04095200035307143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,2560,0.012466666599114736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,2560,0.013907555076811047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,2560,0.01532444523440467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,7168,0.024047111471494038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,2048,0.012475555141766867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,1536,0.010996444357766045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,2048,0.011063111325105032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,3072,0.015388444066047668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,1024,0.008642666869693333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,1536,0.010423111418883005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,1536,0.01313511116637124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,8192,0.025781333446502686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,1024,0.00832799987660514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,1024,0.008346666892369589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,1024,0.011774222056070963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,768,0.00796888851457172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,768,0.0069857777820693115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,512,0.004893333133724001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,768,0.009702222214804756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,2048,0.014512888259357877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,256,0.003604444364706675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,512,0.007294221884674496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,1536,0.009560888840092553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,512,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,512,0.008708444734414419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,256,0.006928000185224745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,2048,0.01311288856797748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,128,0.005636444522274866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,64,0.0028968888024489083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,128,0.00703022215101454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,128,0.007435555259386699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,32,0.0029093333416514923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,64,0.006311111152172089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,65536,0.25851554340786403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,12288,0.03392622205946181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,128,0.0031999999450312722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,256,0.00564444437623024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,65536,0.1527644395828247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,256,0.008016000191370646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,65536,0.13275378280215794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,51200,0.11474400096469456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,51200,0.10625333256191677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,32,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,16384,0.041834665669335254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,12288,0.05753066804673937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,768,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,12288,0.0324924455748664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,16384,0.07275111145443387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,12288,0.035135110219319664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,12288,0.031423999203575984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,10240,0.02810844447877672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,16384,0.044960889551374644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,10240,0.030505776405334473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,8192,0.0383840004603068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,51200,0.11121511459350586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,8192,0.022296888960732356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,7168,0.03291911217901442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,10240,0.04766222172313266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,7168,0.021391110287772283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,16384,0.03839022252294753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,7168,0.023956444528367784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,6144,0.028594666057162817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,51200,0.20285867320166695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,6144,0.01850044396188524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,6144,0.01907822158601549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,5120,0.024713777833514746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,8192,0.027494221925735474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,5120,0.01702222228050232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,5120,0.016364443633291457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,4096,0.02161066730817159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,5120,0.019485332899623446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,7168,0.020311999652120803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,3584,0.018586667047606576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,4096,0.01777777738041348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,8192,0.024373332659403484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,3584,0.015596444408098856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,6144,0.02203111184967889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,3072,0.016433777080641854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,3584,0.013807999591032663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,3072,0.012793777717484368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,3072,0.014146667387750415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,4096,0.014697778556081982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,2560,0.0140364451540841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,65536,0.14306310812632242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,2560,0.011653333074516721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,2560,0.012471111284361945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,2048,0.011759111450778114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,2560,0.01386222243309021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,3584,0.016194666425387066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,1536,0.009742221898502773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,2048,0.011220444407727985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,2048,0.013320888910028668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,1536,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,4096,0.01443822185198466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,3072,0.015372445185979208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,1536,0.012089778151777057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,1024,0.007811555431948767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,1024,0.010326222413116032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,2048,0.010403555300500657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,768,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,512,0.0052382221652401825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,1536,0.008832000195980072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,1024,0.007963555554548899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,10240,0.02604977786540985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,256,0.0038613333470291565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,512,0.007098666495747036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,1024,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,512,0.00868888861603207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,768,0.006796444455782573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,128,0.003248888792263137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,256,0.005578666511509154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,256,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,64,0.002903999967707528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,128,0.005281777845488654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,32,0.0031902222997612427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,768,0.0073315559162033936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,64,0.005962666538026597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,65536,0.23130755954318574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,512,0.0064328888224230866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,65536,0.14397422472635904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,256,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,65536,0.1299040052625868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,51200,0.1799253357781304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,128,0.007820444802443186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,51200,0.10917777485317653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,16384,0.06257600254482694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,768,0.009023110899660323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,51200,0.10157866610421075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,32,0.006270222365856171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,16384,0.040248887406455144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,65536,0.13356799549526638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,16384,0.04098755452367995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,51200,0.10648177729712592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,12288,0.029155555698606703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,128,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,10240,0.02696000039577484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,10240,0.02517866591612498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,8192,0.035511109564039446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,10240,0.028892444239722356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,12288,0.0480657782819536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,16384,0.0385519994629754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,8192,0.025827555192841425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,12288,0.030580444468392268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,10240,0.04113599989149306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,7168,0.020208888583713107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,6144,0.023725334140989516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,12288,0.03301599952909682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,7168,0.02338933282428318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,5120,0.020350222786267597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,6144,0.017889777819315594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,6144,0.021148444877730474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,8192,0.020641777250501845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,5120,0.016686222619480558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,5120,0.018964444597562153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,4096,0.014055111342006259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,7168,0.019468444916937087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,4096,0.017172444197866652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,4096,0.014190221826235453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,6144,0.01777777738041348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,3584,0.013387555877367655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,3072,0.01370933320787218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,4096,0.017180444465743173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,3072,0.01278133359220293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,3072,0.014880888991885714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,3072,0.012079111403889127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,2560,0.011747555600272285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,8192,0.02327644493844774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,2560,0.011183111204041375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,2048,0.01073333372672399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,3584,0.015074667003419666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,2048,0.01037599974208408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,2048,0.00959644466638565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,1536,0.009268444445398118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,2048,0.012888888518015543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,5120,0.016068443655967712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,1536,0.008661333057615492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,1024,0.007673777639865875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,1536,0.011899555722872416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,2560,0.01145511120557785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,1024,0.006981333096822103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,1024,0.009391110804345872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,768,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,3584,0.013763555222087435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,7168,0.02697333362367418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,768,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,512,0.004613333278232151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,2560,0.013946667313575745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,512,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,512,0.008437333007653555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,1024,0.007294221884674496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,3584,0.01625066664483812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,256,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,1536,0.008572444319725037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,256,0.006329777754015393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,256,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,128,0.0052675555149714155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,64,0.003206222214632564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,512,0.006826666494210561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,128,0.007702222300900354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,32,0.003208000005947219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,65536,0.1756124496459961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,32,0.0063440000845326324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,768,0.007613333563009898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,768,0.008957333034939235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,65536,0.09428711069954766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,65536,0.07082755698098077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,128,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,256,0.003563555578390757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,51200,0.07097066773308648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,16384,0.04671466681692335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,51200,0.05747555361853706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,65536,0.13343555397457546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,16384,0.02304977840847439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,128,0.0035582222044467926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,51200,0.10583288802040948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,51200,0.1264479955037435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,12288,0.021684444612926904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,12288,0.018594667315483093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,12288,0.03279644582006667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,10240,0.01905866629547543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,10240,0.01646133263905843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,10240,0.028435554769304063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,64,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,12288,0.03636888994110955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,8192,0.016505777835845947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,16384,0.04057866666052077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,8192,0.024821332759327356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,10240,0.027899556689792212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,7168,0.014595554934607612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,6144,0.018561777141359117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,7168,0.0225591113169988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,8192,0.02401955591307746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,5120,0.01587377819750044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,6144,0.020607110526826646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,16384,0.028392000330819026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,8192,0.014835556348164877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,5120,0.012105777859687805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,5120,0.012072888513406118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,5120,0.018629334039158292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,7168,0.013680000272062091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,4096,0.011036444041464063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,4096,0.016866667403115165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,6144,0.012759110993809171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,3584,0.011062222222487131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,3072,0.011415111521879831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,3584,0.015613334046469795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,7168,0.02094399929046631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,4096,0.01347555551264021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,2560,0.010079999764760336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,3072,0.00925244473748737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,3072,0.014475555883513557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,4096,0.010858666565683154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,2048,0.008348444269763099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,2560,0.008752000000741746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,2560,0.013617777989970313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,2048,0.008375110725561777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,2048,0.01259022206068039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,3072,0.01036711119943195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,1536,0.00699733363257514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,3584,0.011937778029176923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,1536,0.007080000307824876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,2560,0.009528888596428765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,1536,0.010724444356229572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,6144,0.013456000222100152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,768,0.004554666578769684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,1024,0.006679111056857639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,1024,0.0062977779242727495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,1024,0.009375111096435124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,1536,0.007485333416197035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,512,0.003546666767862108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,768,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,768,0.008654221892356873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,256,0.0031999999450312722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,512,0.008011555506123437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,1024,0.005975111077229182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,256,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,128,0.0031928889867332247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,256,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,256,0.007672888537247975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,3584,0.009681777821646797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,128,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,768,0.0063991112013657885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,64,0.0028853333658642242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,32,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,512,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,64,0.00552977787123786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,32,0.005716444303592046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,65536,0.08550488948822021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,2048,0.007873777714040544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,128,0.004938666605287128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,65536,0.06660977999369304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,128,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,512,0.005643555687533484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,16384,0.04412177867359585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,51200,0.054409778780407377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,51200,0.06449421909120348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,16384,0.02621777852376302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,16384,0.021977777282396953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,16384,0.040298667218950056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,51200,0.12184622552659775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,12288,0.019973332683245342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,12288,0.018050667312410142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,10240,0.030172444052166406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,65536,0.13289244969685873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,65536,0.1547057761086358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,8192,0.02216088937388526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,10240,0.017863111363516915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,10240,0.015847111741701763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,51200,0.1059946682718065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,7168,0.018920888503392536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,12288,0.035242666800816856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,8192,0.02455288834042019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,7168,0.013434666726324292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,6144,0.016488000750541687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,7168,0.022264889544910852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,12288,0.032524443335003324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,6144,0.012706666356987424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,6144,0.020607999629444547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,5120,0.01143555591503779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,10240,0.028192000256644353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,8192,0.014369777507252164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,5120,0.01834222177664439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,5120,0.011780444118711682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,7168,0.013912000589900546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,4096,0.010462221999963125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,4096,0.01684177749686771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,5120,0.014474666780895658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,3584,0.010763555765151978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,3584,0.009027555584907532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,3072,0.010348444183667501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,3584,0.015595555305480957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,8192,0.015272888872358533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,4096,0.012744888663291931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,3072,0.010049777726332346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,3072,0.0147724449634552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,4096,0.010034666293197209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,3584,0.01109333336353302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,2560,0.009269333548016017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,2560,0.01368444412946701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,6144,0.012088889049159156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,1536,0.007300444775157505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,2048,0.007637333538797166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,2048,0.012493333054913415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,1536,0.006664888726340399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,3072,0.0086666668454806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,1024,0.005114666703674528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,2560,0.008970666262838576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,1024,0.006614222294754452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,2048,0.008016000191370646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,768,0.0041288890772395665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,1024,0.009015999734401703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,768,0.006303999986913469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,2048,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,512,0.0035502223504914176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,768,0.00871022211180793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,1536,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,512,0.005611555443869696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,512,0.008047999607192146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,1024,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,256,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,128,0.0028657778683635923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,256,0.005980444451173146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,768,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,128,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,128,0.005981333139869902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,512,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,32,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,64,0.005316444569163852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,32,0.005628444254398346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,65536,0.12427200211419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,256,0.007384888827800751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,65536,0.07642489009433322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,65536,0.06407022476196289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,64,0.0029048888633648553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,1536,0.010623999767833285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,51200,0.058168000645107694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,16384,0.0377111103799608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,51200,0.05218933357132805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,2560,0.008336000144481659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,51200,0.10574933555391099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,16384,0.021151999632517498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,12288,0.030390222867329914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,128,0.007975999679830339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,256,0.0031991110493739447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,12288,0.0184515549076928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,65536,0.13312710656060114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,12288,0.017549332645204332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,12288,0.031932443380355835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,10240,0.016157334049542744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,8192,0.022330666581789654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,10240,0.027894222074084814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,16384,0.04007733199331496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,51200,0.09887555572721694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,8192,0.01388177772363027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,7168,0.019783111082182992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,8192,0.024107555548350017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,10240,0.026505778233210247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,6144,0.017916444275114272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,7168,0.013037333057986366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,7168,0.021918222308158875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,10240,0.01554844445652432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,5120,0.015999111864301894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,6144,0.020008888509538438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,5120,0.010760000182522668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,5120,0.018230222993426852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,4096,0.011904000408119626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,7168,0.01295555548535453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,4096,0.009921777579519484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,3584,0.010509333676762052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,6144,0.012054222325483957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,3584,0.010202666951550378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,8192,0.014544889330863953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,3072,0.009367999931176504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,3584,0.008991111483838823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,5120,0.011223999990357293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,3072,0.008361777497662438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,2560,0.00832622249921163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,4096,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,6144,0.01200355589389801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,2560,0.008538666698667739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,2560,0.007667555577225155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,2560,0.013201778133710226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,3584,0.015376889043384127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,2048,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,2048,0.007359999749395583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,2048,0.011528000235557556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,3072,0.009720889230569204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,3072,0.01447111037042406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,1536,0.006659555352396435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,4096,0.016171556380059984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,1536,0.01036888857682546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,2048,0.007567110988828871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,1024,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,1024,0.006046222315894232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,1024,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,768,0.006266666783226862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,1536,0.007024889190991719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,512,0.0035946667194366455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,768,0.008719111482302347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,512,0.006270222365856171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,512,0.005583999885453119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,16384,0.023929778072569106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,512,0.008009778128729926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,256,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,256,0.0052373334765434265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,256,0.007668444679843054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,768,0.0058240000572469496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,128,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,64,0.0028453332682450614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,128,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,32,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,1024,0.005245333330498801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,1536,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,64,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,65536,0.1094942225350274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,32,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,768,0.004222222086456087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,128,0.0032088889016045462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,65536,0.06950310866038005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,65536,0.06189333067999946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,16384,0.03291022115283542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,51200,0.05028533273273044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,65536,0.13302577866448298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,128,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,12288,0.026760889424218073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,16384,0.020303999384244282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,16384,0.03990222348107232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,12288,0.016758221719000075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,12288,0.01705511079894172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,10240,0.023323555787404377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,12288,0.03199999862247043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,10240,0.015145777000321282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,51200,0.052720887793434985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,10240,0.02751111156410641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,51200,0.10542222526338364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,16384,0.021735999319288466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,8192,0.013800889253616333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,7168,0.017887110511461895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,8192,0.013401777380042605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,8192,0.023698666029506262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,7168,0.012441777520709567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,7168,0.012101333174440594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,7168,0.021732444564501446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,10240,0.015079999963442484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,8192,0.02031822171476152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,6144,0.011360000405046674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,5120,0.014490667316648694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,6144,0.011429333024554782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,6144,0.020045333438449435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,4096,0.010755555497275459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,256,0.0032151111712058387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,5120,0.010218666659461128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,5120,0.01800355480776893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,3584,0.009907555249002244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,4096,0.009400889277458191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,51200,0.08818222416771783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,6144,0.01643822259373135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,4096,0.016179554992251925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,3072,0.008647110727098253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,3584,0.00833155545923445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,3584,0.009810667071077559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,2560,0.007979555262459649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,3584,0.015199111567603217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,3072,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,3072,0.014118222726715935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,2560,0.007640000018808577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,2560,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,5120,0.010207999911573198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,2560,0.01310400002532535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,1536,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,4096,0.00869955536392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,2048,0.006992888947327931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,2048,0.011528888510333167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,1024,0.004567999806669024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,1536,0.006953777538405524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,1536,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,1536,0.010567111273606619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,768,0.004237333519591225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,1024,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,1024,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,1024,0.009022222624884712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,768,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,768,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,768,0.00870666652917862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,256,0.003236444460021125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,512,0.0052844443255000645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,512,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,256,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,256,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,128,0.0029235554652081597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,512,0.0039022221333450745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,2048,0.006687110910813014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,128,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,64,0.0028799999919202593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,128,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,128,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,2048,0.007354666789372762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,65536,0.08860266870922512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,64,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,32,0.005431111074156231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,51200,0.0711324413617452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,256,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,65536,0.13227111763424343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,51200,0.04833689000871447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,16384,0.026288888520664636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,51200,0.048093332184685596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,51200,0.10532710949579875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,16384,0.01924444403913286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,12288,0.021028444170951843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,65536,0.06269244352976482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,3072,0.008866666919655269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,12288,0.015490666031837463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,10240,0.01889333294497596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,12288,0.03162311183081733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,32,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,512,0.006040888941950268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,10240,0.014510222607188754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,8192,0.01647822227742937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,10240,0.027408889598316614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,16384,0.03967466619279649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,7168,0.014375110467274984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,8192,0.023737778266270954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,12288,0.01644088824590047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,7168,0.01144977741771274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,6144,0.013462222284740873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,10240,0.013979555832015144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,16384,0.01960711181163788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,6144,0.010715555813577441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,5120,0.01240533341964086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,65536,0.05977333254284329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,6144,0.01993777851263682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,5120,0.009357333183288574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,4096,0.010832000109884473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,7168,0.021661332911915247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,4096,0.008842666943868002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,4096,0.016168889072206285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,8192,0.0129120002190272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,5120,0.009687110781669617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,3584,0.008736888567606607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,7168,0.011788444386588203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,3584,0.008028444316652086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,3072,0.009092444347010719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,4096,0.008364444805516137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,8192,0.013096888860066732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,3072,0.007733333441946242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,2560,0.008314666648705801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,5120,0.017895999881956313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,2560,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,3584,0.014803555276658801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,2560,0.012128888732857175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,2560,0.007486222518814935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,6144,0.010441777606805166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,2048,0.006573333508438534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,1536,0.005016889009210798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,2048,0.010738666686746808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,3072,0.013780444032616086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,1536,0.00665155549844106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,1024,0.004212444441186057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,1536,0.010023111270533668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,2048,0.006160000132189856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,1024,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,1024,0.006445333361625671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,3584,0.00963199966483646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,1024,0.009010666774378883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,2048,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,768,0.006269333263238271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,512,0.0035573335157500375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,768,0.008691555923885768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,256,0.003208000005947219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,512,0.005961777849329843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,512,0.005296000176005893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,512,0.00812266684240765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,128,0.002895111011134254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,256,0.005883555445406172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,256,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,768,0.003563555578390757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,3072,0.007682667010360294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,64,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,32,0.0028773333049482773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,128,0.007359999749395583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,768,0.005413333161009683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,64,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,32,0.005247999810510212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,1536,0.006254222244024277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,51200,0.06103733513090345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,65536,0.05978577666812473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,65536,0.0587315559387207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,256,0.005221333354711533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,16384,0.02427111069361369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,51200,0.04662400152948168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,128,0.005281777845488654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,51200,0.10505511363347371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,16384,0.019681778219011094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,12288,0.019544000426928203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,16384,0.03919733232922024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,12288,0.0151555554734336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,10240,0.017204445269372728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,12288,0.031189332405726116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,65536,0.1329288880030314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,51200,0.047205332252714366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,10240,0.01348088930050532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,8192,0.015035554766654968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,10240,0.027103111147880554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,16384,0.018566222654448617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,65536,0.07532266775767009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,8192,0.012469333079126147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,8192,0.02369511127471924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,12288,0.016172443826993305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,7168,0.011398221883508893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,6144,0.01314755529165268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,7168,0.02188088827662998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,6144,0.010418666733635796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,128,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,10240,0.014507555299335055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,8192,0.01276622215906779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,5120,0.009431111315886179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,5120,0.01754755609565311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,5120,0.009688888986905416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,7168,0.011471110913488599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,4096,0.008781332936551835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,4096,0.008350222474998897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,3584,0.00960000024901496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,6144,0.010392000277837118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,5120,0.012110222544935016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,3072,0.008667555948098501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,3584,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,3584,0.014519999424616495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,7168,0.0136817776494556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,2560,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,3072,0.007965332931942409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,3072,0.01312711089849472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,2560,0.006812444577614467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,2560,0.0074746666683091064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,2048,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,2560,0.012089778151777057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,4096,0.015955555770132277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,1536,0.004884444591071871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,2048,0.007183999651008182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,2048,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,3584,0.007896000312434303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,1536,0.006731555693679386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,1024,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,1536,0.010070222119490305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,1024,0.005628444254398346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,1024,0.009404444032245213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,1024,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,6144,0.01958933307064904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,2048,0.011062222222487131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,768,0.0052675555149714155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,1536,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,768,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,768,0.00872444444232517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,512,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,256,0.002971555623743269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,512,0.008160889148712158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,768,0.003926222109132343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,256,0.005149333427349727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,256,0.005690666536490123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,256,0.007930667036109501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,512,0.003548444559176763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,128,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,64,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,3072,0.007385777930418651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,128,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,512,0.005223999834722943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,64,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,32,0.0053671110007498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,65536,0.05688889159096611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,128,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,65536,0.13281155957116023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,4096,0.010363555616802638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,51200,0.04480888777308994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,16384,0.02122666603989071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,65536,0.06678844160503812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,51200,0.1052853332625495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,16384,0.0192586663696501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,16384,0.018174222773975797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,12288,0.01754755609565311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,65536,0.057814220587412514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,16384,0.039072000318103366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,10240,0.015508444772826301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,12288,0.0158906661801868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,12288,0.031136890252431233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,128,0.004951999833186467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,8192,0.01347733371787601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,10240,0.02750844425625271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,51200,0.04623111089070638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,8192,0.011935999823941125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,7168,0.012499555945396423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,8192,0.023937778340445623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,51200,0.051664001411861844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,12288,0.014576888746685453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,7168,0.010953777366214328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,6144,0.011415999796655444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,32,0.0029031110720502008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,10240,0.014133334159851074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,5120,0.010646222366227044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,6144,0.010065777434243096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,6144,0.0196088883611891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,8192,0.012170666621790992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,5120,0.009342222577995723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,4096,0.010392888552612728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,5120,0.017416889468828838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,4096,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,4096,0.015776000089115567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,10240,0.013111111190583972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,6144,0.010031110710567897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,3584,0.007300444775157505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,3072,0.008364444805516137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,5120,0.009248889154858058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,3584,0.014479999740918478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,3072,0.007680888805124495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,4096,0.008378666308191087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,3584,0.00869777798652649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,7168,0.02168888847033183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,2560,0.006613333192136552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,2560,0.007367110914654202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,2048,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,2560,0.012367111113336353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,2048,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,2048,0.006329777754015393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,2048,0.011071110765139261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,3072,0.007015110717879401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,2560,0.0069546666410234236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,1024,0.003875555677546395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,1536,0.006770666688680649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,1536,0.00977955593003167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,1024,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,1024,0.009384888741705153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,3584,0.008061333662933773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,1536,0.004571555389298333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,768,0.005535999933878581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,768,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,1536,0.006001777946949005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,512,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,1024,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,512,0.008348444269763099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,256,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,256,0.005144000053405762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,256,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,256,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,128,0.002918222298224767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,128,0.004968889057636261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,512,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,128,0.005962666538026597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,64,0.002556444456179937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,128,0.007827555967701806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,32,0.002577777744995223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,512,0.0032435556252797446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,65536,0.05945510996712578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,3072,0.01312444441848331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,64,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,32,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,51200,0.04667288727230496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,65536,0.05528888768619961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,768,0.0038817777401871155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,65536,0.13252445062001547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,16384,0.01942044496536255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,51200,0.04311466548177931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,51200,0.04553066690762838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,12288,0.016542222764756944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,16384,0.018939556346999276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,16384,0.017468444175190397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,7168,0.010716444088353051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,16384,0.03905422157711453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,10240,0.014169777433077494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,12288,0.015229332778188916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,65536,0.05651377969317966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,10240,0.013236444857385425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,10240,0.01311644415060679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,10240,0.027095110880004034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,8192,0.011756444142924415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,51200,0.10535733567343818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,768,0.00871288859181934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,12288,0.014132445057233175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,12288,0.031253334548738264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,8192,0.012239110966523489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,6144,0.010497777826256223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,6144,0.009391999906963771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,6144,0.019564444820086162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,5120,0.009701333112186855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,7168,0.010703110860453712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,5120,0.008911111288600499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,7168,0.010415111151006486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,4096,0.008848889006508721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,5120,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,5120,0.01755466726091173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,8192,0.011424889167149862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,3584,0.008594666918118795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,6144,0.009975999593734741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,3584,0.007324444750944774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,3584,0.013847110999955071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,3584,0.008040889269775814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,7168,0.02131822208563487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,3072,0.006948444578382704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,7168,0.011708444191349877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,2560,0.006659555352396435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,2560,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,3072,0.007254222200976477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,2560,0.007636444436179266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,2560,0.01185155577129788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,2048,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,4096,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,8192,0.02401688860522376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,1536,0.004562666846646203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,2048,0.006256888724035687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,2048,0.011061333119869232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,1536,0.0069217777086628815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,1024,0.0041786668201287585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,1536,0.010025777750545079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,1024,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,1024,0.009358222285906473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,768,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,1024,0.0064088888466358185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,3072,0.007694222033023834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,768,0.006167999986145232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,768,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,768,0.008648888932334052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,512,0.005240889059172736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,512,0.008271999657154083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,256,0.003206222214632564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,4096,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,4096,0.01516711049609714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,2048,0.00720000018676122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,128,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,256,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,256,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,512,0.0035457776652442086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,1536,0.005968889014588461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,64,0.0028479999552170434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,128,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,128,0.007300444775157505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,65536,0.04877600073814392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,64,0.0052968888647026485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,32,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,512,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,51200,0.03757866554790073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,3072,0.013075555364290873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,65536,0.05249866512086657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,65536,0.05549777878655327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,256,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,65536,0.13251288731892905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,51200,0.041491554843054876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,51200,0.1050106684366862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,128,0.004964444372389051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,12288,0.0136817776494556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,16384,0.017052445146772597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,16384,0.018479110466109384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,16384,0.03899022274547153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,10240,0.011739555332395764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,12288,0.01475822263293796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,12288,0.014171555638313293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,12288,0.031119111511442397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,10240,0.013099555340078143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,10240,0.0270471109284295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,8192,0.010964444114102257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,32,0.0025395556456512874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,7168,0.012285333540704517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,7168,0.010053333308961656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,8192,0.012891555825869242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,51200,0.04513066675927904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,7168,0.010445333189434476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,7168,0.021548445026079815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,8192,0.023192000057962205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,5120,0.010444444086816577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,5120,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,16384,0.016524444023768108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,6144,0.011853333148691388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,4096,0.009703111317422655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,5120,0.009043555292818282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,8192,0.01109333336353302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,6144,0.019095111224386428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,4096,0.008371555142932469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,10240,0.012701333396964602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,3584,0.007328000333574083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,3584,0.008100444244013892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,3072,0.006640000061856375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,5120,0.01687111126052009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,6144,0.009914666414260864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,4096,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,3072,0.007816000117195977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,2560,0.006530666516886816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,3584,0.013799111048380533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,2560,0.012079999678664737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,6144,0.009078222016493479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,2048,0.0052435555391841466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,3072,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,2560,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,1536,0.004577777865860197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,2048,0.007290666302045186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,2048,0.01108000013563368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,1536,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,1536,0.010077333284748925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,2560,0.007605333295133378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,3584,0.00868800034125646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,2048,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,768,0.00360622215602133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,1536,0.006739555547634761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,1024,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,768,0.00564444437623024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,512,0.0032462223122517266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,768,0.008665777742862701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,4096,0.014768888552983602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,512,0.005314666777849197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,256,0.0028968888024489083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,1024,0.006265777680608962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,512,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,256,0.0057760001056724125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,256,0.004962666581074397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,256,0.0076142218377855085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,1024,0.00562755556570159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,128,0.007336888876226213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,128,0.00591822216908137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,64,0.0028880000528362063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,32,0.0028560000161329904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,1024,0.009024000002278222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,65536,0.03747289048300849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,64,0.004939555707905027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,32,0.004938666605287128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,128,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,51200,0.029498666524887085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,65536,0.050329777929517955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,65536,0.05493244528770447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,768,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,16384,0.01348355578051673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,51200,0.0443742209010654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,51200,0.03978666663169861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,51200,0.10495111015107896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,16384,0.01756266587310367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,16384,0.016501333978441026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,16384,0.03873599900139703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,12288,0.013803555733627744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,10240,0.018240888913472492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,128,0.005072000126043956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,10240,0.011918221910794577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,12288,0.011976888610257043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,10240,0.027130666706297133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,8192,0.01475822263293796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,12288,0.013479111095269522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,8192,0.010575999816258749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,12288,0.03090755475891961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,7168,0.013800000150998434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,8192,0.0107342220014996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,65536,0.13256355126698813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,3072,0.012817777693271637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,7168,0.010413332945770687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,7168,0.020272889071040683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,10240,0.012429333395428128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,6144,0.008845333423879411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,5120,0.011424889167149862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,6144,0.01829600003030565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,8192,0.022667555345429316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,5120,0.008167111211352879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,4096,0.009729777773221334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,5120,0.016407110624843173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,7168,0.009360888765917884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,512,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,4096,0.00833777752187517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,4096,0.014783999986118741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,6144,0.009670221971140968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,3584,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,3584,0.01347555551264021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,5120,0.00905777762333552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,3072,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,6144,0.01240888900227017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,3072,0.007666666474607255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,4096,0.007401777638329401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,2560,0.006587555425034628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,2560,0.00741244438621733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,2560,0.012086222569147745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,3584,0.007320889168315464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,2048,0.006980444822046492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,2048,0.011047999891969891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,2560,0.005657777604129579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,3072,0.01276533305644989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,3584,0.009008000294367472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,1024,0.0038746665749284956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,2048,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,3072,0.006534222099516127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,1024,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,1024,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,2048,0.006260444306664997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,1536,0.004574222283230888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,768,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,768,0.005321777943107817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,512,0.0031911111954185697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,1536,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,1536,0.010046222143703038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,512,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,512,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,768,0.0035546666218174826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,256,0.005236444373925527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,1024,0.009015111459626092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,256,0.007714666426181793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,1536,0.006643555644485686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,128,0.004948444250557157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,64,0.002528888897763358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,256,0.0031617778456873367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,128,0.005650666852792104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,32,0.0026791110220882627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,512,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,65536,0.026887999640570745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,64,0.004950222041871813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,32,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,51200,0.021920000513394673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,65536,0.048410667313469775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,65536,0.1316186719470554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,768,0.00869688888390859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,51200,0.038839111725489296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,16384,0.01312444441848331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,128,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,16384,0.01646488904953003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,12288,0.011387555963463254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,16384,0.038570665650897555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,128,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,65536,0.054770665036307446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,10240,0.010053333308961656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,12288,0.013471999930010902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,12288,0.03055822186999851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,10240,0.01110400011142095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,51200,0.10444888803693984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,10240,0.026428444517983332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,16384,0.015519110692871941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,8192,0.009763555394278632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,8192,0.01074488874938753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,7168,0.00851288851764467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,8192,0.022504000200165644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,12288,0.013059555656380124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,7168,0.01036977767944336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,7168,0.009374221993817223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,7168,0.020131554868486192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,10240,0.012129777835475074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,51200,0.04393511017163595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,6144,0.008708444734414419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,6144,0.01796533332930671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,5120,0.008685333033402761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,4096,0.007622222105662028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,256,0.005702222386995952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,6144,0.008037333687146505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,4096,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,4096,0.008366222182909647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,4096,0.014801777071423002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,5120,0.00815822184085846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,5120,0.016514667206340365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,3584,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,3584,0.013775111072593264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,3072,0.0063288890653186375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,8192,0.009777777724795872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,3072,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,3072,0.01275199982855055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,2560,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,3584,0.007289778027269576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,2560,0.007332444190979004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,2048,0.005655111124118169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,2560,0.011741333537631564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,2048,0.006265777680608962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,3584,0.007344000041484833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,1536,0.004902222090297275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,2048,0.01106844428512785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,3072,0.007663999994595845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,1024,0.00423200014564726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,1536,0.00665511108107037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,1536,0.009834667046864828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,2560,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,768,0.003867555409669876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,1024,0.006305777778228124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,1024,0.009508444203270806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,768,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,2048,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,512,0.003528000166018804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,512,0.00526577772365676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,512,0.008359111017651027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,256,0.0032035555276605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,1024,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,256,0.005106666849719154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,768,0.006160888820886612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,128,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,1536,0.006039111150635614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,128,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,512,0.005619555711746216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,128,0.00566133318675889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,64,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,32,0.002895111011134254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,128,0.007687999970383114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,768,65536,0.022409778502252366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,64,0.004906666775544484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,32,0.004959110998445087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,65536,0.0476399991247389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,65536,0.05411644445525276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,65536,0.1314622163772583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,51200,0.03813955518934462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,51200,0.043729778793123036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,768,16384,0.014479999740918478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,256,0.005633777628342311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,768,51200,0.018654222289721172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,16384,0.03868800070550706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,16384,0.01620533400111728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,768,12288,0.00981600003110038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,51200,0.10416266653272842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,5120,0.007710222568776872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,12288,0.013443555268976422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,16384,0.014831110835075378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,10240,0.011062222222487131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,10240,0.02593066626124912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,10240,0.012127111355463663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,768,8192,0.008839111361238692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,6144,0.009551111194822524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,12288,0.012448000411192576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,768,10240,0.009506666825877296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,8192,0.022397332721286353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,8192,0.010729778144094678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,7168,0.010039110978444418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,7168,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,7168,0.020230222079488967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,6144,0.008722667064931657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,6144,0.018256889449225534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,6144,0.009687999884287516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,8192,0.009850666754775578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,768,7168,0.00832444429397583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,5120,0.009016888837019602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,768,4096,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,5120,0.01651733285850949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,4096,0.007419555551475949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,4096,0.014831110835075378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,768,3584,0.0069857777820693115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,768,5120,0.007276444799370236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,12288,0.02985066506597731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,3584,0.008026666939258575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,5120,0.008119111259778341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,3584,0.013627555635240344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,768,6144,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,3072,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,3072,0.0070062221752272705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,768,2560,0.00666844430896971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,3072,0.012949333422713809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,4096,0.008365333080291748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,2560,0.006330666856633292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,2560,0.011575111084514193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,3584,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,768,3072,0.00683111117945777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,2048,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,768,1536,0.004566222015354368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,1536,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,2560,0.007263999846246507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,768,2048,0.004946666873163647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,768,1024,0.003949333396222857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,1536,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,1536,0.010086221827401055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,2048,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,768,768,0.003556444413132138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,1024,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,2048,0.010667555862002902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,1024,0.009005332986513773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,768,512,0.0032088889016045462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,768,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,768,0.005296000176005893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,768,0.008786666724416945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,768,256,0.0032071111102898917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,256,0.007669332954618666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,512,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,256,0.005624889085690181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,256,0.004968889057636261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,256,0.008000888758235509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,1024,0.005640888793600931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,128,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,128,0.007677333222495184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,768,32,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,64,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,512,65536,0.02884977724817064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,512,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,32,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,512,51200,0.024029332730505202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,65536,0.053492443429099194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,65536,0.046966221597459584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,65536,0.13147911760542128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,768,64,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,51200,0.10427466366026138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,16384,0.014856000741322836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,512,0.006181333214044571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,512,12288,0.009193778038024902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,16384,0.0377075539694892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,768,128,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,12288,0.012464000119103326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,512,10240,0.008089777496125963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,12288,0.029676444000667993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,128,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,10240,0.012082666986518435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,10240,0.011201777391963534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,10240,0.025847999585999384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,16384,0.016226665841208566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,8192,0.009823111196359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,512,7168,0.008047111332416534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,8192,0.010842666857772403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,12288,0.013460444079505073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,51200,0.04061155517896017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,512,6144,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,7168,0.02032444377740224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,512,8192,0.008274666965007782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,6144,0.008767999708652496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,6144,0.009708444277445475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,512,5120,0.007130666739410824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,6144,0.018586667047606576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,8192,0.022517333428064983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,5120,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,5120,0.016547555724779766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,51200,0.0380328893661499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,4096,0.0073315559162033936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,4096,0.008286221987671321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,4096,0.01482311056719886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,3584,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,3584,0.013806222213639153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,5120,0.009062222308582729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,512,4096,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,512,16384,0.009708444277445475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,3072,0.00702311098575592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,512,2560,0.006664888726340399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,512,3584,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,2560,0.006659555352396435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,2560,0.012071111136012606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,512,2048,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,3584,0.008012444608741337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,7168,0.010060444474220276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,2048,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,512,1536,0.004566222015354368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,3072,0.01275644451379776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,1536,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,1536,0.010030222435792288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,1536,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,2560,0.007371555599901412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,512,3072,0.006960888703664144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,1024,0.006275555739800136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,1024,0.005611555443869696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,1024,0.009054222040706212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,2048,0.011092444260915121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,3072,0.007598222129874759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,512,512,0.0032142222755485107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,768,0.006190222170617845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,768,0.00871377769443724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,512,0.004984000076850255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,512,1024,0.0038862224254343244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,512,0.008255110846625434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,7168,0.009391110804345872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,256,0.004974222017659081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,2048,0.006955555743641323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,512,128,0.002867555452717675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,768,0.0053031109273433685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,256,0.007991111112965478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,128,0.005642666584915585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,512,64,0.0025519999778932994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,512,256,0.0030942221896515954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,512,32,0.002516444358560774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,512,768,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,256,65536,0.015511110424995422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,32,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,256,0.005618666609128316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,65536,0.044738666878806226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,256,51200,0.013571555415789286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,512,0.006010666489601135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,128,0.007734222544564142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,51200,0.037709332174725003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,64,0.004963555683692296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,256,16384,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,51200,0.10284711254967584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,16384,0.01605688863330417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,16384,0.014799111419253878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,16384,0.037535998556349016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,65536,0.04689066608746847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,256,10240,0.007634667058785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,12288,0.012441777520709567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,51200,0.03782755467626784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,12288,0.029879109727011785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,256,8192,0.007335999773608313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,10240,0.011120888921949597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,10240,0.025812443759706285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,8192,0.010079999764760336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,256,7168,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,8192,0.02197066611713833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,256,12288,0.00833422193924586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,7168,0.010044444766309526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,12288,0.013444444371594323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,7168,0.020256888535287645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,6144,0.008674666285514832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,6144,0.018591110905011494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,6144,0.009701333112186855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,8192,0.010790222220950656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,10240,0.012098666694429187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,5120,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,128,0.004570666700601578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,256,6144,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,4096,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,4096,0.014486221803559197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,256,3584,0.006314666734801398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,256,5120,0.006008000009589725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,3584,0.007152888509962294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,5120,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,65536,0.12975645065307617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,5120,0.016539555456903245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,3072,0.006802666518423293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,3072,0.012929778132173749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,4096,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,256,2560,0.006337777607970768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,3072,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,2560,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,3584,0.013904889424641928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,7168,0.009348444640636444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,3584,0.007648000286685095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,2560,0.01182044463025199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,256,1536,0.004598222259018156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,2048,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,2048,0.0110871113008923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,1536,0.006018666757477655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,256,1024,0.0037066667444176147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,256,2048,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,256,4096,0.005978666659858491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,1024,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,256,768,0.003564444267087512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,1024,0.009011555876996782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,1536,0.006642666541867786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,768,0.005384888913896348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,256,512,0.003245333416594399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,768,0.008039111064540016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,2560,0.007277333074145847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,256,3072,0.006292444550328785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,256,256,0.0031555555760860443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,512,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,2048,0.00630044440428416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,256,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,256,0.005649777750174205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,256,0.007663110891977946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,1536,0.009700444837411245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,768,0.006109333286682765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,128,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,128,0.007672000262472365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,1024,0.005629333357016246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,256,32,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,128,65536,0.01458666721979777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,512,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,32,0.004974222017659081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,65536,0.04348266786999173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,256,128,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,65536,0.1289573378033108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,128,51200,0.012772444221708508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,128,0.005607999861240387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,51200,0.03467822074890137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,512,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,128,16384,0.008001777860853408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,51200,0.10203288661109076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,64,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,128,12288,0.007648888561460707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,16384,0.01584533353646596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,16384,0.01461866663561927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,16384,0.037018666664759316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,128,10240,0.007295110987292395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,12288,0.012423111332787408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,12288,0.02983022067281935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,256,64,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,10240,0.011071110765139261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,128,8192,0.006776000062624614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,10240,0.025748444928063288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,51200,0.03742311067051358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,128,7168,0.006665777828958299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,8192,0.010805333654085795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,8192,0.009728000395827824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,8192,0.02236088944805993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,128,6144,0.006331555545330048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,7168,0.009104000197516548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,65536,0.046329776446024575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,6144,0.009745777481132084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,128,5120,0.006040888941950268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,6144,0.018255111243989732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,5120,0.008042666647169325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,5120,0.008980444735950893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,128,4096,0.005998222364319696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,5120,0.016183111402723525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,7168,0.010057777994208867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,4096,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,4096,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,7168,0.02028622229894002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,4096,0.01462399959564209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,128,3072,0.005990222096443176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,3584,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,10240,0.012108444339699216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,3072,0.007648888561460707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,128,2560,0.0063386667105886675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,3072,0.012794666820102267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,128,3584,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,12288,0.013429332938459186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,128,2048,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,2560,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,6144,0.008725333544943068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,3584,0.014010666145218743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,2048,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,128,1536,0.004565333326657613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,2048,0.01074488874938753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,3072,0.007032000356250339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,128,1024,0.003923555629120933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,1536,0.006416888700591193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,1536,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,1536,0.01038755559258991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,128,768,0.003559111307064692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,1024,0.005667555663320754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,2560,0.011765333513418833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,3584,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,768,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,128,512,0.0031795555518733132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,768,0.00848444468445248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,2048,0.007002666592597961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,128,256,0.0028924445311228433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,512,0.005987555616431766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,512,0.004943999979231092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,512,0.008303999900817871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,128,128,0.0028444443725877335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,256,0.005622222191757626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,1024,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,2560,0.006344888773229387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,128,64,0.002528000002106031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,1024,0.00871377769443724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,128,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,128,32,0.002521777732504739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,128,0.007151111132568783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,768,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,64,51200,0.010046222143703038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,64,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,64,16384,0.007683555285135905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,65536,0.04317155480384827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,64,12288,0.006355555521117316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,51200,0.034782220919926964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,64,10240,0.006524444454246097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,16384,0.014487110906177096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,64,8192,0.0069431112044387394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,256,0.007715555528799693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,12288,0.012121777567598553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,64,7168,0.006360888895061281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,128,0.004895110925038655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,64,6144,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,64,5120,0.006321777900060018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,64,65536,0.011182222101423474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,64,4096,0.005394666559166378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,7168,0.009039999710188972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,6144,0.00868444475862715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,64,3584,0.005974222388532426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,5120,0.008183110919263627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,64,3072,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,4096,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,64,2560,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,256,0.004968889057636261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,64,2048,0.005977777971161737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,3072,0.006996444529957241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,10240,0.011067555182509951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,64,1536,0.004259555704063839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,2560,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,64,1024,0.003913777983850903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,64,768,0.003524444583389494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,1536,0.005988444305128521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,1024,0.005631111148330901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,768,0.005294222384691238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,64,256,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,3584,0.007251555720965068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,64,128,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,256,0.005250666704442766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,64,64,0.002542222125662698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,2048,0.0063235556913746735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,64,32,0.002536000063021978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,32,0.004902222090297275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,64,512,0.003273777870668305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,32,0.004954666727119022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,32,51200,0.008700444466537898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,512,0.005033777819739448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,65536,0.04316177633073595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,32,16384,0.006400888992680444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,32,12288,0.006160000132189856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,51200,0.03501155641343858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,8192,0.009742221898502773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,32,8192,0.005595555735958948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,10240,0.011067555182509951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,32,7168,0.006330666856633292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,8192,0.009715555442704095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,32,6144,0.005997333261701796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,7168,0.009404444032245213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,32,5120,0.005997333261701796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,32,10240,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,32,4096,0.005668444352017508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,128,0.004945777770545748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,32,3584,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,64,0.004910222358173794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,4096,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,3584,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,32,2560,0.006268444574541516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,3072,0.0069306666652361555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,32,2048,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,2560,0.0063279999627007385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,6144,0.008639111287064022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,16384,0.014490667316648694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,12288,0.012077333198653327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,1536,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,32,3072,0.00589155571328269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,32,768,0.0035573335157500375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,32,512,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,1024,0.005387555393907759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,32,256,0.0026391111314296722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,768,0.005492444253630108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,32,128,0.002567111204067866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,512,0.0052728888889153796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,32,64,0.002536000063021978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,256,0.004934222333961063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,32,1536,0.004553777890072929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,32,32,0.002509333400262727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,5120,0.008092444803979661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,32,1024,0.003975999852021535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,64,0.004591111093759537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,32,32,65536,0.009396444592210982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,16384,0.36153422461615664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,16384,0.16378222571478948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,51200,0.48893245061238605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,51200,1.125937779744466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,2048,0.006310222049554189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,16384,0.13678399721781412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,32,0.00490399988161193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,12288,0.12210222085316975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,10240,0.10362578100628322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,51200,0.4416666560702854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,12288,0.10367111365000407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,10240,0.10427555772993301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,8192,0.1851644383536445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,7168,0.1621333360671997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,8192,0.08761155605316162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,8192,0.0714577767584059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,128,0.004908444566859139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,7168,0.0786844425731235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,5120,0.11576178338792588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,6144,0.0885653363333808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,4096,0.09379822015762329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,5120,0.07311555412080553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,4096,0.04003644320699904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,4096,0.05980977747175428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,3584,0.08198222186830309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,12288,0.26692356003655326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,3072,0.07139644357893202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,3584,0.05192710955937704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,6144,0.056178669134775795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,10240,0.22324800491333008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,5120,0.047227554851108126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,3072,0.031593776411480375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,2560,0.05965244770050049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,2048,0.04804089003139072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,2560,0.03917510973082648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,6144,0.1390568945142958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,1536,0.03802488909827338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,3584,0.03560444381501939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,2048,0.03386844529045953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,1024,0.02640177806218465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,1536,0.022405332989162866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,1536,0.024638222323523626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,3072,0.046310222811169095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,768,0.020657777786254883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,1024,0.01921777758333418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,512,0.014302222265137566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,2560,0.027760000692473516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,768,0.015623999966515435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,256,0.00907022257645925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,512,0.012085333466529846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,2048,0.023355555203225877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,128,0.006324444380071428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,256,0.010627555350462595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,256,0.009176000124878353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,64,0.004200889004601372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,128,0.0107351111041175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,128,0.008031110796663497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,1024,0.01555022266176012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,64,0.010388444695207808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,768,0.013801777528391944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,32,0.010017777482668558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,7168,0.0925991137822469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,512,0.011805333197116852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,16384,0.2819555600484212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,65536,0.5175929069519043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,32,0.004596444467703502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,51200,0.39119556215074325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,65536,1.0891218185424805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,51200,0.855987548828125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,12288,0.2077955537372165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,16384,0.13144532839457193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,10240,0.1769733296500312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,12288,0.08322044213612874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,51200,0.34003021981981063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,8192,0.1435991128285726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,10240,0.08365777466032241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,8192,0.056932442718082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,7168,0.0721342232492235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,7168,0.05066133207745022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,12288,0.09847644302580093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,65536,0.45965157614813906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,10240,0.08685421943664551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,6144,0.04504799842834473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,6144,0.06552710798051622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,5120,0.09112177954779731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,4096,0.07412888606389363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,5120,0.04650133185916477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,5120,0.058443552917904325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,3584,0.0656871133380466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,4096,0.03261244297027588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,8192,0.0710808899667528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,3072,0.05685955286026001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,6144,0.10659911235173543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,3584,0.04311733444531759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,2560,0.04790488878885905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,3072,0.02608977754910787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,3072,0.03743199838532342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,2048,0.03854488995340135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,2560,0.03243466549449497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,2048,0.01944977707333035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,1536,0.030735111898846094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,2048,0.028115553988350764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,1024,0.021968000464969214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,1536,0.016524444023768108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,4096,0.04636977778540718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,7168,0.1262213389078776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,768,0.016542222764756944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,1024,0.013552889227867126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,512,0.01203555530971951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,16384,0.12827022870381674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,768,0.012495111260149213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,256,0.007673777639865875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,768,0.013608888619475894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,512,0.010784889260927836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,512,0.010727110836240979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,256,0.009701333112186855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,2560,0.02330044408639272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,256,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,128,0.009691555466916826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,32,0.004272888931963178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,1536,0.02134399943881565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,64,0.009009777671760982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,32,0.008845333423879411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,65536,0.3332364559173584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,65536,0.17484800020853677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,128,0.004564444637960858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,51200,0.2605680094824897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,64,0.0038746665749284956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,65536,0.157843550046285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,16384,0.08407466941409641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,128,0.007350222104125553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,51200,0.1318355533811781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,12288,0.06260266568925646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,51200,0.12421866257985432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,3584,0.029454221328099568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,12288,0.03655022382736206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,12288,0.033519112401538424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,8192,0.04473333226309883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,10240,0.03587199913130866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,10240,0.03177777926127116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,7168,0.038450668255488075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,8192,0.024548444483015273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,8192,0.02751022246148851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,6144,0.033414221472210355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,7168,0.022264889544910852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,16384,0.048242668310801186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,7168,0.030682666434182063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,1024,0.015525332755512662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,10240,0.05277155505286323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,4096,0.025027554896142747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,5120,0.018032888571421306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,5120,0.023457777169015672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,3584,0.02162222233083513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,4096,0.015903110305468243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,3072,0.018942221999168396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,3584,0.014531556102964612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,3584,0.01830222209294637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,2560,0.01626666718059116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,16384,0.047298666503694325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,2560,0.012085333466529846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,2560,0.01479022204875946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,4096,0.020543111695183646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,2048,0.012851555314328937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,2048,0.011414222419261932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,1536,0.010798222488827176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,1536,0.009691555466916826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,3072,0.013457777599493662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,1536,0.011400889191362592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,1024,0.00833155545923445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,768,0.007112888826264276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,768,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,768,0.007841777470376756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,512,0.004941333499219683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,6144,0.020093333390023973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,512,0.007338666253619724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,256,0.003565333369705412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,2048,0.013423999978436364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,256,0.0069466663731469055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,256,0.005617777920431561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,128,0.005566222386227713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,5120,0.028927998410330877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,64,0.0029155556112527847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,128,0.0069822221994400024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,32,0.003216000066863166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,1024,0.008793777889675563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,64,0.006324444380071428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,128,0.0032471112079090546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,65536,0.14669777287377253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,6144,0.02659911082850562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,51200,0.20099022653367785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,512,0.006330666856633292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,51200,0.11273333761427139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,16384,0.07219555642869738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,51200,0.10628266466988458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,65536,0.255058659447564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,12288,0.05723110834757487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,16384,0.041999111572901406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,16384,0.04179822074042427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,10240,0.04759644468625387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,12288,0.030043555630577937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,3072,0.016770665844281513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,10240,0.03010933266745673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,8192,0.02238133384121789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,7168,0.03541422221395705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,32,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,7168,0.026224889689021643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,6144,0.03126044405831231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,1024,0.008398222426573435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,10240,0.027785778045654297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,6144,0.018231110440360177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,5120,0.027433777848879497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,8192,0.02423200011253357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,4096,0.02067733307679494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,5120,0.020779555042584736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,4096,0.01458577811717987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,7168,0.020206222931543987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,65536,0.1325368881225586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,3584,0.016557332542207506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,3584,0.013456888496875763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,5120,0.01625422139962514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,3072,0.015221332510312399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,3072,0.012456888953844706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,4096,0.01847822301917606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,3584,0.017640888690948486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,2048,0.011713777979214987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,2560,0.011456888582971362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,2048,0.00996888925631841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,3072,0.015977778368526034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,2560,0.013532444834709167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,2560,0.013693333499961428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,1536,0.008692444198661381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,1024,0.007841777470376756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,1024,0.007744000189834171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,1024,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,768,0.006578666468461354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,1536,0.010227555202113258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,768,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,768,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,512,0.006259555617968242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,512,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,256,0.0035582222044467926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,8192,0.03994755612479316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,128,0.003531555748648114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,256,0.006859555426571105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,6144,0.025621333056026038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,512,0.004648000001907349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,64,0.0032142222755485107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,32,0.0031964443624019623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,64,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,32,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,65536,0.2185120052761502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,12288,0.031818668047587075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,65536,0.14147555828094482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,256,0.005623111294375525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,128,0.0069226668112807805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,65536,0.1263777812321981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,16384,0.062423110008239746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,51200,0.10236622227562799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,51200,0.107859558529324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,16384,0.03967466619279649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,12288,0.030232889784706965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,12288,0.03070666723781162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,10240,0.042287998729281955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,51200,0.1714320050345527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,10240,0.028063111835055884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,10240,0.026688888669013977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,128,0.005646222167544895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,7168,0.03190222382545471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,8192,0.021533333592944678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,16384,0.038703110482957624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,12288,0.04875555634498596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,6144,0.0281333327293396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,7168,0.019647111495335896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,7168,0.02272355556488037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,5120,0.024127110838890076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,6144,0.01754488878779941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,4096,0.018560889694425795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,6144,0.019955555597941082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,5120,0.017720000611411203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,5120,0.01757777730623881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,2048,0.01200177768866221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,8192,0.03642755415704515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,3584,0.013096888860066732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,3072,0.01424800025092231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,3584,0.014521777629852295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,8192,0.022964444425370958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,3072,0.011589333415031433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,3072,0.013023110727469126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,2048,0.010868444210953183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,2560,0.011973333027627734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,3584,0.0159582214223014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,1536,0.009666666388511658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,4096,0.015864888827006023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,1536,0.00832622249921163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,1536,0.00905511114332411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,1024,0.007596444752481248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,1024,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,768,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,2048,0.009389333426952362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,4096,0.013963555296262106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,512,0.004584889031118817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,768,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,512,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,256,0.0038888889054457345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,2560,0.010381333529949188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,1024,0.007335111498832703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,256,0.005586666779385672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,128,0.003272888975010978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,2048,0.010407999985747868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,128,0.006797333144479328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,64,0.0031884445084465873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,128,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,32,0.00319377767542998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,64,0.005969777703285217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,65536,0.17635111014048258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,512,0.006073777874310811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,51200,0.13915110958947077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,65536,0.09093688593970405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,256,0.006367110957702001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,16384,0.04274310999446445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,1536,0.009294222626421187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,51200,0.058752887778811984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,12288,0.03304622239536709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,16384,0.024033778243594702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,2560,0.012460444536474017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,12288,0.01924444403913286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,10240,0.02801333202256097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,10240,0.018554665976100497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,10240,0.01901333365175459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,8192,0.023788443870014612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,51200,0.0694755580690172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,8192,0.01479644411140018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,16384,0.027950220637851294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,7168,0.02081155611409081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,8192,0.016528889536857605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,32,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,12288,0.02163022259871165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,7168,0.017329777280489605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,6144,0.012687111066447364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,65536,0.07025333245595296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,6144,0.01685244507259793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,4096,0.014124444789356656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,5120,0.012712889247470431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,768,0.00666844430896971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,5120,0.01422933406300015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,3584,0.012294222083356647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,4096,0.010634666515721215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,3072,0.011058666639857821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,3584,0.009678222239017487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,3584,0.012116444607575735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,6144,0.01851022243499756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,2560,0.009658666948477427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,7168,0.01371911085314221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,3072,0.01109955542617374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,2560,0.008363555702898238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,5120,0.016152888536453247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,2560,0.00998488896422916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,2048,0.007686222592989604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,2048,0.00904888908068339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,1024,0.005592888842026393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,1536,0.007646222081449296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,1536,0.007365333537260692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,768,0.00423200014564726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,4096,0.0131217779384719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,1024,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,768,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,768,0.006106666806671355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,2048,0.008671110702885522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,512,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,512,0.005640888793600931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,128,0.0028764444092909494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,256,0.005924444645643234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,128,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,64,0.0028799999919202593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,1024,0.006656000183688269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,32,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,3072,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,256,0.003208000005947219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,32,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,1536,0.007484444313579136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,256,0.005241777747869492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,51200,0.11985511249966091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,65536,0.08402222394943237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,128,0.004936000125275718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,51200,0.06314044528537326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,512,0.0036968886852264404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,16384,0.02586311101913452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,12288,0.034232000509897866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,16384,0.021877333521842957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,10240,0.02951644526587592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,65536,0.1535386641820272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,12288,0.019855999284320407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,8192,0.021920000513394673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,10240,0.016116445263226826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,10240,0.017582222819328308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,65536,0.06655466556549072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,16384,0.04435644547144572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,8192,0.015494222442309061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,51200,0.05383822321891785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,7168,0.016882666283183627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,64,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,6144,0.01240799989965227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,6144,0.015903110305468243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,12288,0.01792977750301361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,4096,0.013204444613721637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,5120,0.01108266661564509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,5120,0.013415999710559845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,3584,0.011154666543006897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,4096,0.010020444790522257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,4096,0.012111110819710625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,3072,0.010061333576838175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,6144,0.01696622206105126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,3584,0.011429333024554782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,2560,0.009027555584907532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,3072,0.008709333009190029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,3072,0.010713777608341642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,2048,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,2560,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,2560,0.009744000103738572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,1536,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,2048,0.007432888779375289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,2048,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,7168,0.019162666466501024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,8192,0.014121777481502958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,1536,0.0070088886552386815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,3584,0.008982222113344405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,768,0.003906666818592283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,7168,0.013220444321632385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,1024,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,768,0.006310222049554189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,768,0.005963555640644497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,512,0.005965333431959152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,256,0.0032231110251612137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,1024,0.004607111215591431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,256,0.005296000176005893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,128,0.0032195556494924757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,1024,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,512,0.003552888830502828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,128,0.006089777582221561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,64,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,128,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,512,0.005624889085690181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,64,0.0052435555391841466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,32,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,1536,0.006959999601046245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,65536,0.07723289065890841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,256,0.006016888966162999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,5120,0.014511111709806653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,51200,0.05705955293443468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,16384,0.03784711162249247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,16384,0.0210826661851671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,16384,0.023522666758961145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,51200,0.10097066561381023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,65536,0.06349422534306844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,12288,0.01826844447188907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,12288,0.01755644381046295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,8192,0.022037333912319605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,51200,0.05174044436878628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,32,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,7168,0.02035288843843672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,8192,0.013786666923099093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,8192,0.01447111037042406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,6144,0.01758666667673323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,7168,0.012881778180599213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,7168,0.015291555060280694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,5120,0.0158595558669832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,6144,0.011920000116030375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,6144,0.015079999963442484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,10240,0.025976000560654536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,4096,0.012920888761679331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,5120,0.012421333127551608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,10240,0.015609777635998197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,3584,0.010467555787828235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,10240,0.01606399979856279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,4096,0.011672000090281168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,3584,0.008847999903890822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,3584,0.010840000377760993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,2560,0.008658666577604082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,3072,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,3072,0.010045333041085137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,2048,0.0075777777367168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,2560,0.007679111427730984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,5120,0.010682666467295753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,1536,0.006320000108745363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,2048,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,4096,0.009462222456932068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,1024,0.004455111092991299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,1536,0.00685777763525645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,1536,0.006983111302057902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,768,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,1024,0.006610666712125142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,65536,0.1268328825632731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,1024,0.006261333409282897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,768,0.006268444574541516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,768,0.005997333261701796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,256,0.0031866667171319327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,2560,0.009015999734401703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,256,0.005975111077229182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,128,0.002903999967707528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,3072,0.009647111097971598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,128,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,128,0.004960888789759742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,512,0.0035217776894569397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,32,0.002895111011134254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,512,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,512,0.005585777676767773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,65536,0.10919377538892959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,2048,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,65536,0.0680604444609748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,12288,0.030214223596784804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,51200,0.05163288778728909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,16384,0.03219199842876858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,64,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,32,0.0052933332820733385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,16384,0.020319110817379422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,12288,0.026151999831199646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,256,0.0052391112678580815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,10240,0.022827555735905964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,65536,0.0614337788687812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,12288,0.016514667206340365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,8192,0.020061333974202473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,64,0.0028693332440323303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,10240,0.014861333701345654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,7168,0.01773422294192844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,51200,0.049916442897584706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,8192,0.013728000223636627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,6144,0.016024000114864774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,7168,0.012433778080675336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,7168,0.014555555250909595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,16384,0.021391999390390184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,5120,0.014283556077215405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,6144,0.011420444481902652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,51200,0.08608711428112453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,4096,0.011060444845093621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,5120,0.010030222435792288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,10240,0.01514311134815216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,5120,0.012143111063374413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,3584,0.009679999616410997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,4096,0.009031111167536842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,8192,0.013307555682129331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,3072,0.00868622213602066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,3584,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,2560,0.00796977761718962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,3072,0.007918222082985772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,3072,0.009371555513805812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,2048,0.006952889263629913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,2560,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,6144,0.013167111410035027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,12288,0.016890666551060148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,2048,0.007320889168315464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,2048,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,1024,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,1536,0.0069848886794514125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,1536,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,4096,0.011039999624093374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,1024,0.0063786668082078295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,1024,0.005981333139869902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,512,0.0038862224254343244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,768,0.0063075555695427795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,768,0.005778666585683823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,256,0.0031893334041039148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,512,0.00599911105301645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,512,0.0052800000541739995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,1536,0.006260444306664997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,128,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,256,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,256,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,64,0.002867555452717675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,128,0.00573244442542394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,128,0.004909333255555895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,768,0.0042568888101312844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,64,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,32,0.005286222116814719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,51200,0.07087555858823988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,65536,0.059234665499793164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,65536,0.06224089198642307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,2560,0.007401777638329401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,16384,0.02667644454373254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,51200,0.04725777771737841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,16384,0.019664888580640156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,16384,0.019031110737058852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,10240,0.018622222873899672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,12288,0.015233778291278414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,10240,0.014474666780895658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,3584,0.01036977767944336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,51200,0.04786755641301473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,8192,0.012543111211723752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,7168,0.01401511165830824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,8192,0.01291822228166792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,12288,0.016224000189039443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,7168,0.011743110915025076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,6144,0.012440889245933957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,7168,0.013614222407341003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,5120,0.01127466642194324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,65536,0.08956088622411092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,6144,0.01055466632048289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,10240,0.013817778064144982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,4096,0.010382222632567087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,5120,0.00906399968597624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,3584,0.009022222624884712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,5120,0.01109955542617374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,4096,0.010372444159454769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,3584,0.007982222570313348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,3584,0.009198221895429824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,32,0.0029048888633648553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,3072,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,8192,0.01622222198380364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,6144,0.013730666703648038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,2048,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,2560,0.007329777710967594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,4096,0.008387555678685507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,1536,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,12288,0.02113866640461816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,2048,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,3072,0.008225777910815345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,1024,0.0038933331767717996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,1536,0.006704888823959563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,768,0.003567111161020067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,1024,0.006309333360857434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,2560,0.007287110719415877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,1024,0.006002666635645761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,768,0.006024888820118374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,768,0.005616888817813661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,2560,0.007368000017272101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,512,0.005380444642570283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,256,0.005980444451173146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,128,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,2048,0.007070221834712558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,1536,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,128,0.004957333207130432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,512,0.0032071111102898917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,512,0.0058337777025169795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,64,0.005057777795526716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,256,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,32,0.005039111193683413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,3072,0.008004444340864817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,64,0.00254844439526399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,65536,0.058509336577521436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,51200,0.061567107836405434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,65536,0.058431108792622886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,256,0.0032186667538351486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,16384,0.025077333052953083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,51200,0.046929776668548584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,51200,0.04526577724350823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,12288,0.01997599999109904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,16384,0.01924622224436866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,10240,0.01686755485004849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,12288,0.016201777590645682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,12288,0.014866666661368476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,8192,0.014922666880819531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,10240,0.013468444347381592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,128,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,7168,0.013473778135246701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,8192,0.012426666915416718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,8192,0.012390221986505719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,6144,0.01184800018866857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,7168,0.011404444773991903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,7168,0.013213333156373767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,6144,0.010035555395815108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,6144,0.013471110827393003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,5120,0.009210666848553551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,65536,0.07721155881881714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,5120,0.011095110740926532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,3584,0.008492444124486711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,10240,0.014488889111412896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,4096,0.009407111340098912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,3072,0.007807110746701558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,3584,0.008050666915045844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,3584,0.008839999636014303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,2560,0.006940444310506185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,3072,0.00794311116139094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,32,0.0028880000528362063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,3072,0.007660444411966536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,2560,0.0073626670572492815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,2560,0.007702222300900354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,4096,0.009701333112186855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,2048,0.006736889067623351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,4096,0.008560889297061497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,1536,0.006674666785531574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,1536,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,1024,0.006405333264006509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,768,0.003531555748648114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,5120,0.010703999963071613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,16384,0.018603555030292936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,768,0.005614222337802251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,512,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,512,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,256,0.0032195556494924757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,2048,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,768,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,256,0.005146666533417172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,128,0.0028453332682450614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,128,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,64,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,1024,0.004049777984619141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,32,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,1024,0.00583022211988767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,65536,0.06775555345747206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,256,0.00563466673096021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,1536,0.004898666507667965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,65536,0.055861334005991616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,51200,0.05285155442025927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,65536,0.057224002149369985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,128,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,51200,0.04610844453175863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,51200,0.043599112166298755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,12288,0.018238221605618794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,16384,0.019208888212839764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,16384,0.017671111557218764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,32,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,12288,0.0158906661801868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,12288,0.014529777897728814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,2048,0.0070017774899800616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,10240,0.013148444394270578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,8192,0.011954666839705573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,7168,0.01277955538696713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,64,0.004991111242108875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,10240,0.015832889411184523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,7168,0.01054666688044866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,512,0.003546666767862108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,6144,0.011750222080283694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,16384,0.021293333835071985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,5120,0.010785777535703448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,5120,0.009012444151772393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,4096,0.00998222248421775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,7168,0.011767999993430244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,4096,0.008365333080291748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,3584,0.008548444343937768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,6144,0.01147466649611791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,3584,0.0074755557709270054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,3584,0.008055110772450766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,3072,0.008414222134484185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,3072,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,8192,0.013648889131016202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,2560,0.007648000286685095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,3072,0.007747555772463481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,2560,0.00739822205570009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,2048,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,2048,0.006413333531883027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,2048,0.0070293330483966405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,10240,0.014136888914638095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,1536,0.006673777682913675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,1536,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,1024,0.004262222184075249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,2560,0.006828444285525217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,768,0.004052444464630551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,1024,0.006333333336644703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,1024,0.005614222337802251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,512,0.0038746665749284956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,768,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,4096,0.008202667037645975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,512,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,6144,0.010034666293197209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,256,0.005912000106440649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,128,0.0028693332440323303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,768,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,128,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,1536,0.005503111001518037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,512,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,5120,0.009638222555319468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,32,0.002571555475393931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,65536,0.06038577689064873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,64,0.005038222091065513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,256,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,32,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,51200,0.04717688759167989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,65536,0.05644444624582926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,65536,0.05346666773160299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,16384,0.019293333093325298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,128,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,51200,0.042041778564453125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,12288,0.014198222094111972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,12288,0.015192889504962497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,10240,0.013993778162532382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,12288,0.01605422298113505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,10240,0.01348622226052814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,8192,0.012443555725945367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,8192,0.011397333608733283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,8192,0.011090666883521609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,7168,0.011530666715568967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,16384,0.018936889039145578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,6144,0.01070666644308302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,7168,0.010414222048388587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,7168,0.011075555450386472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,5120,0.010087110930018956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,6144,0.00980444418059455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,5120,0.00906755526860555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,4096,0.008941333327028487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,16384,0.01719377769364251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,4096,0.00758844448460473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,3584,0.00796977761718962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,4096,0.00831200016869439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,6144,0.009728888670603434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,3072,0.00721155520942476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,5120,0.008671110702885522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,3072,0.00700800038046307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,2560,0.00664533343580034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,3072,0.007972444097201029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,64,0.0028657778683635923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,2048,0.005671111245950063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,2560,0.006680000159475539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,3584,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,2048,0.0070222218831380205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,3584,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,1536,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,1024,0.004535111288229625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,256,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,1024,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,768,0.003927111211750242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,1024,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,768,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,2048,0.006309333360857434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,1536,0.00535999983549118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,512,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,1536,0.005976000179847081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,512,0.005290666802061929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,51200,0.04523644513554043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,2560,0.0074622225430276655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,128,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,128,0.005774222314357758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,64,0.002577777744995223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,10240,0.012799999780125089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,256,0.005010666532648934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,64,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,65536,0.048960887723498873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,32,0.004957333207130432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,512,0.003544888976547453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,65536,0.05554222067197164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,256,0.0031999999450312722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,256,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,128,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,16384,0.018556444181336295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,51200,0.04490844408671061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,16384,0.018575999471876357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,768,0.005375999957323074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,16384,0.016743999388482835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,65536,0.050685332881079785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,12288,0.014488889111412896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,10240,0.013745778136783175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,10240,0.012508444488048553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,10240,0.012426666915416718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,32,0.002530666689078013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,8192,0.010839111275143094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,8192,0.011045333411958484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,12288,0.014867555763986377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,7168,0.010423111418883005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,12288,0.013841778039932251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,8192,0.012059555285506778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,5120,0.009666666388511658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,6144,0.009692444569534725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,7168,0.011412444214026133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,5120,0.009173333644866943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,4096,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,5120,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,3584,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,4096,0.008067555725574493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,7168,0.010059555371602377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,6144,0.010763555765151978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,4096,0.007431999676757389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,3072,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,3584,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,3072,0.00703022215101454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,2560,0.007359999749395583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,2048,0.005097777893145879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,2560,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,2048,0.007095110913117726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,51200,0.03831111060248481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,3584,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,1536,0.005972444597217772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,51200,0.04047822289996677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,1024,0.003927999900446998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,1536,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,768,0.0038746665749284956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,1024,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,1024,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,512,0.003527111063400904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,768,0.006275555739800136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,768,0.0052791109515561005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,256,0.0032222221295038858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,1536,0.004594666676388847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,512,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,128,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,256,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,3072,0.007754666937722101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,6144,0.00868088917599784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,64,0.0028693332440323303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,128,0.005617777920431561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,128,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,65536,0.03788622220357259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,64,0.004967111266321606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,32,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,51200,0.030215111043718126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,2048,0.006335111127959357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,256,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,51200,0.044450667169359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,51200,0.03888355692227682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,2560,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,12288,0.011685333318180509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,16384,0.017331555485725403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,16384,0.016400888562202454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,12288,0.013458666702111563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,12288,0.013455111119482251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,65536,0.055086221959855824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,10240,0.012259555359681448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,16384,0.013536888692114087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,8192,0.010037333601050908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,7168,0.011390222443474663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,32,0.0028702221396896574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,7168,0.010415111151006486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,6144,0.01072266697883606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,512,0.005977777971161737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,6144,0.008832000195980072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,5120,0.009696889254781935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,8192,0.011066666907734342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,5120,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,4096,0.00885333369175593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,10240,0.010810666614108615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,7168,0.009341333475377824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,3584,0.007768889268239339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,4096,0.0074444446298811175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,65536,0.04905600017971463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,3072,0.006267555471923616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,3584,0.007984889050324758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,3584,0.007294221884674496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,2560,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,3072,0.007763555480374231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,3072,0.0069982219073507525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,2560,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,2560,0.006613333192136552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,5120,0.00904977818330129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,4096,0.008369777765538957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,2048,0.006269333263238271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,8192,0.014818666709793938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,6144,0.009735999835862054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,1536,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,1024,0.005611555443869696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,1024,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,768,0.0035217776894569397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,2048,0.00703111125363244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,10240,0.011759111450778114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,512,0.00332622230052948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,768,0.006245333287451003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,1536,0.006717333363162146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,1024,0.003925333420435588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,512,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,512,0.005283555636803309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,128,0.0028560000161329904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,2048,0.0052435555391841466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,128,0.00565422202150027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,128,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,1536,0.004548444516128964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,32,0.002918222298224767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,64,0.0048942222363419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,256,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,32,0.005110222018427319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,51200,0.024069334069887798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,65536,0.05450755688879225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,65536,0.0470071103837755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,256,0.004958222309748332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,16384,0.012387555506494312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,51200,0.043997334109412305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,51200,0.03798222210672166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,16384,0.016175111134847004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,16384,0.015853333804342482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,10240,0.010746666954623329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,12288,0.012470222181744046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,65536,0.02752888864941067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,8192,0.009123555488056606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,10240,0.012125333150227865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,10240,0.011113777756690979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,7168,0.008816888762844933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,256,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,8192,0.009705777797434065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,6144,0.00869599978129069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,7168,0.01039644413524204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,12288,0.011742221812407175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,64,0.0028444443725877335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,6144,0.009687110781669617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,6144,0.00868444475862715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,4096,0.007305777735180325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,5120,0.009030222064918941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,5120,0.008065777520338694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,3584,0.006960888703664144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,4096,0.008364444805516137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,4096,0.007452444069915348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,3072,0.006266666783226862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,3584,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,3072,0.007637333538797166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,5120,0.007976888782448238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,12288,0.013356444736321768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,768,0.005299555758635203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,2560,0.007157333195209503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,2048,0.005586666779385672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,2048,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,8192,0.010809777511490716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,2048,0.006204444501135085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,1536,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,1024,0.004219555606444676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,1024,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,1024,0.005967999911970562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,768,0.0038986665507157645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,3584,0.007967111137178209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,768,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,7168,0.009355555805895064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,768,0.005274666680230035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,2560,0.005575110928879843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,256,0.003195555466744635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,512,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,512,0.0053191110491752625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,2560,0.006605333338181178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,256,0.005102222164471944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,128,0.00565155554148886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,64,0.0025271111064487034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,512,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,3072,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,32,0.002879111096262932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,768,65536,0.023771555887328252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,32,0.004947555561860402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,65536,0.04645333356327481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,65536,0.05443288882573446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,128,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,51200,0.043848001294665866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,51200,0.03734044565094842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,768,16384,0.011517333487669626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,1536,0.005972444597217772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,768,12288,0.01110666659143236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,16384,0.016005333926942613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,16384,0.014666666587193808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,128,0.0029102222373088202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,768,10240,0.010021333065297868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,12288,0.013251555462678274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,768,51200,0.019501333435376484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,1536,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,768,8192,0.00870666652917862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,10240,0.011077333655622272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,768,7168,0.008039111064540016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,8192,0.01072799993885888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,64,0.004953777624501122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,8192,0.009694221946928237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,768,6144,0.008627555436558193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,7168,0.010417777631017896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,768,5120,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,6144,0.009658666948477427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,256,0.005626666463083691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,6144,0.008762666748629676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,768,4096,0.007650666766696506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,5120,0.008277333445019191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,768,3584,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,4096,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,3584,0.008022222254011366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,768,3072,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,7168,0.009376889301670922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,3072,0.007682667010360294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,3072,0.0069511110583941145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,4096,0.008010666403505538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,10240,0.012084444363911947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,2560,0.006445333361625671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,2560,0.00721066693464915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,2048,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,768,1536,0.004606222112973531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,2048,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,768,1024,0.0038817777401871155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,1536,0.006641777853171031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,12288,0.012479110724396177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,768,768,0.003584889074166616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,1024,0.006010666489601135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,768,2048,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,768,0.005315555466545953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,3584,0.007293333609898885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,512,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,512,0.0053173332578606075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,768,256,0.003165333428316646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,1536,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,768,128,0.002830222249031067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,768,2560,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,256,0.005743999862008625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,768,64,0.0025759999536805684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,256,0.004955555415815777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,128,0.005639111002286275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,768,32,0.0028693332440323303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,128,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,768,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,768,512,0.0032888888898822996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,64,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,32,0.004960000101062986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,512,51200,0.016174222032229107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,512,16384,0.01036977767944336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,65536,0.045913779073291354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,51200,0.03699288765589396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,512,12288,0.010061333576838175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,5120,0.009007111191749573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,16384,0.014513777361975776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,1024,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,512,65536,0.018135999639829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,10240,0.011051555474599203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,512,8192,0.00832622249921163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,65536,0.04934044347869026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,8192,0.009912889036867354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,8192,0.0107342220014996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,16384,0.016031111280123394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,512,10240,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,512,6144,0.008314666648705801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,7168,0.00925866680012809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,6144,0.009425777528021071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,512,5120,0.008007110820876228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,51200,0.038507554266187884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,5120,0.007648000286685095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,5120,0.008711111214425828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,512,4096,0.007318221860461765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,12288,0.013456000222100152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,512,3584,0.006654222392373615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,4096,0.007337777978844113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,3584,0.008018666671382057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,512,3072,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,3072,0.00665244460105896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,512,7168,0.009045333498054082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,512,2560,0.005587555468082428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,3072,0.007672888537247975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,7168,0.010173333187898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,512,2048,0.004951999833186467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,2560,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,2560,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,512,1536,0.004579555657174853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,6144,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,1536,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,1536,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,512,1024,0.003871110992299186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,1024,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,12288,0.012192000117566852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,1024,0.006375999914275275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,2048,0.006960888703664144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,3584,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,512,512,0.0032017777363459268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,768,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,512,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,512,256,0.002918222298224767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,512,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,4096,0.008082666330867344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,512,128,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,256,0.004907555464241239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,512,768,0.0035413333939181436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,512,64,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,10240,0.012128888732857175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,512,32,0.002558222247494592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,768,0.006148444281684027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,256,65536,0.015896000795894198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,64,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,32,0.005012444323963589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,256,51200,0.015108444624476962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,65536,0.04648355642954508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,65536,0.04333066609170702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,256,0.005745777653323279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,256,16384,0.009566222627957663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,51200,0.0373671121067471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,51200,0.03497066762712266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,16384,0.015899555550681222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,128,0.005676444619894028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,256,10240,0.008658666577604082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,2048,0.006112888869312074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,12288,0.013168888787428537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,12288,0.012426666915416718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,10240,0.011744000017642975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,256,7168,0.007695111135641734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,8192,0.010798222488827176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,8192,0.00997866690158844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,256,6144,0.007281777759393056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,7168,0.010370666782061258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,7168,0.009019555317031013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,256,5120,0.00674222202764617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,6144,0.00976533359951443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,256,12288,0.009735999835862054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,6144,0.008839111361238692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,256,4096,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,16384,0.014486221803559197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,5120,0.008004444340864817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,256,3584,0.0069520001610120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,4096,0.00833155545923445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,4096,0.007360888852013483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,10240,0.011083555718262991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,3584,0.007141333487298753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,256,2560,0.006675555474228329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,3072,0.00666222224632899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,2560,0.007326222128338284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,2560,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,5120,0.00907733374171787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,128,0.0046453335218959385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,256,8192,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,2048,0.005977777971161737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,256,3072,0.006680000159475539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,256,1024,0.0036435553597079385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,1536,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,1536,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,256,768,0.0035502223504914176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,1024,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,256,2048,0.005980444451173146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,768,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,256,1536,0.004585777719815572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,512,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,256,256,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,3072,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,1024,0.006275555739800136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,256,128,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,256,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,768,0.006253333141406377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,256,512,0.0032346666687064697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,128,0.004594666676388847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,256,32,0.002583111118939188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,2048,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,128,65536,0.016916443904240925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,3584,0.007979555262459649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,32,0.004569777597983678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,128,51200,0.014120888378885059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,65536,0.04606133368280199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,65536,0.04271733429696825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,128,16384,0.00868711123863856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,51200,0.033951110310024686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,128,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,256,64,0.0026311110705137253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,16384,0.015850666496488783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,16384,0.01456622282663981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,128,10240,0.007679111427730984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,12288,0.01309866706530253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,64,0.004931555440028508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,12288,0.012084444363911947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,512,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,10240,0.011758222348160215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,10240,0.011100444528791638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,8192,0.010417777631017896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,8192,0.009692444569534725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,51200,0.037343111303117536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,128,6144,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,7168,0.010078222387366826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,128,12288,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,128,5120,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,6144,0.009500444763236577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,6144,0.008742222355471717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,128,4096,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,128,8192,0.0070257774657673305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,128,7168,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,4096,0.008367111285527548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,128,3584,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,4096,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,256,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,3584,0.007869333028793335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,3584,0.007318221860461765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,128,2560,0.005972444597217772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,3072,0.00664444433318244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,2560,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,5120,0.008960888617568547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,5120,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,2048,0.0070773329999711775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,128,1536,0.00423555572827657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,128,3072,0.005919111271699269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,1536,0.006647999915811751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,128,1024,0.00388711111413108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,1024,0.0063484443558586975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,128,2048,0.005919999960396025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,2048,0.006262222097979651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,768,0.006066666709052191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,7168,0.009366222553782994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,1536,0.005924444645643234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,512,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,128,256,0.0028826666788922418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,1024,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,256,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,256,0.004940444396601783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,768,0.005315555466545953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,128,0.005648889061477449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,128,0.004933333231343163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,3072,0.007432888779375289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,128,32,0.0025395556456512874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,512,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,2560,0.006609777609507243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,32,0.004615111069546806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,64,51200,0.01091644416252772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,128,128,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,64,16384,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,51200,0.03394044438997904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,64,12288,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,16384,0.014469333820872836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,64,10240,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,64,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,128,768,0.003567999849716822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,10240,0.011065777805116443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,64,7168,0.006607111129495833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,8192,0.009743111001120673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,64,6144,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,7168,0.009169778062237633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,128,512,0.0032231110251612137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,64,5120,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,64,4096,0.005707555760939916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,6144,0.008685333033402761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,64,3584,0.006008000009589725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,12288,0.012087110843923358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,5120,0.008110221889283922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,64,3072,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,4096,0.007295110987292395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,64,2560,0.005658666706747479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,64,2048,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,3072,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,65536,0.0423351095782386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,64,1536,0.004591111093759537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,64,1024,0.0036159998012913596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,128,64,0.0028488888508743713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,64,768,0.003519111209445529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,1024,0.005587555468082428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,64,512,0.0031964443624019623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,64,65536,0.012771555946932899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,64,256,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,3584,0.007172444628344641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,64,128,0.0025395556456512874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,2560,0.00638488887084855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,256,0.004942222187916438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,2048,0.005998222364319696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,64,32,0.002518222149875429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,1536,0.005991999887757831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,32,65536,0.01312800000111262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,32,0.004587555511130227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,32,51200,0.010751999914646149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,768,0.0052897776994440295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,32,16384,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,65536,0.042451557185914784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,512,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,51200,0.03393777873780992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,32,10240,0.0069573331210348345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,128,0.00498488876554701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,12288,0.012101333174440594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,10240,0.010727110836240979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,64,0.004876444323195351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,32,7168,0.006249777972698212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,64,8192,0.006790222393141852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,32,6144,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,8192,0.009701333112186855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,32,5120,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,7168,0.009187555147541894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,32,4096,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,6144,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,32,3584,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,5120,0.007888889147175683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,32,8192,0.006682666639486949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,4096,0.007328000333574083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,32,2560,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,3584,0.006950221955776215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,32,2048,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,3072,0.00702133360836241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,32,1536,0.004606222112973531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,2560,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,32,1024,0.0038942222793896994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,32,12288,0.007677333222495184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,32,768,0.003532444437344869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,1536,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,64,64,0.0025111111915773815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,32,512,0.0032435556252797446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,1024,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,32,3072,0.005704000178310607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,32,128,0.002528000002106031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,512,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,32,64,0.0028417776856157514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,128,0.004583111239804162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,32,32,0.002528888897763358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,64,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,2048,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,16384,0.014487110906177096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,17,32,256,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,768,0.005371555685997009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,256,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,51200,0.4888746473524306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,51200,0.9717555575900607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,16384,0.16362222035725912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,12288,0.2272124555375841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,16384,0.13730043835110134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,12288,0.12212444676293267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,51200,0.42707022031148273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,12288,0.1042284435696072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,16384,0.32795288827684194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,10240,0.10325600041283502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,10240,0.09491822454664443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,32,0.004578666554556953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,16384,0.2827271090613471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,10240,0.1890746619966295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,8192,0.08811644713083903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,8192,0.07226577732298109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,7168,0.13416266441345215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,10240,0.1755546728769938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,7168,0.09091911051008437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,8192,0.13946133189731175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,8192,0.15565155612097845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,7168,0.0655040012465583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,6144,0.0786693361070421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,6144,0.05549600058131748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,7168,0.1428791152106391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,12288,0.20655377705891928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,6144,0.12309688991970485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,4096,0.08030133114920722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,6144,0.11540799670749241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,5120,0.11434666315714519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,4096,0.05628533495797051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,3584,0.069541335105896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,4096,0.039954665634367205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,5120,0.09735466374291314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,4096,0.08551555871963501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,3072,0.060290667745802135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,3584,0.035387555758158364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,3584,0.07515644364886813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,3072,0.04666399955749512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,2560,0.050864888562096484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,3072,0.0640231106016371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,51200,0.879136880238851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,2560,0.02760800056987339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,2048,0.04279822111129761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,3584,0.04958666695488823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,2048,0.03302844365437826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,1536,0.03238755464553833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,2048,0.023752000596788194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,5120,0.047351999415291675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,1536,0.02604444490538703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,1536,0.02233155568440755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,1024,0.023226666781637404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,1536,0.03898310992452834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,2560,0.061371558242373996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,768,0.018897778458065458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,1024,0.015836444165971544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,1024,0.018909333480728995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,1024,0.03378933336999681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,512,0.013618666264745923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,768,0.015146666102939181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,3072,0.03142044279310439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,768,0.028130667077170476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,512,0.012454222473833295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,512,0.012562666502263812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,512,0.02271822260485755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,256,0.009684444301658207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,256,0.018572444717089336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,128,0.006550222221348021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,2048,0.04998311069276598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,5120,0.07068533367580838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,128,0.008246222303973304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,64,0.005913777897755305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,128,0.010829333629873065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,2560,0.03943822118971083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,128,0.01554844445652432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,64,0.010207999911573198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,32,0.01038933296998342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,256,0.010675555302037133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,768,0.013735999663670858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,65536,0.5171288914150661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,65536,1.0984257592095268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,32,0.0058462222417195635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,51200,0.8486622174580892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,51200,0.3910355567932129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,16384,0.28161777390374076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,65536,0.44352266523573136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,256,0.008021333151393467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,16384,0.1311875581741333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,51200,0.3443875577714708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,12288,0.20793155829111734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,10240,0.17302132977379692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,12288,0.08257599671681722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,10240,0.0833360022968716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,16384,0.2317884498172336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,10240,0.07693422502941556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,8192,0.14139555560217962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,12288,0.09830488761266072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,8192,0.07086400190989177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,7168,0.12297511100769043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,8192,0.05820888943142361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,8192,0.11848178174760605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,7168,0.07310577895906237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,12288,0.17282133632236055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,6144,0.10644000106387669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,7168,0.05140711201561821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,10240,0.14334755473666722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,6144,0.06413244538837008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,5120,0.09072355429331462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,6144,0.09612622525956895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,5120,0.05778844488991631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,5120,0.09327644771999782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,5120,0.039112889104419284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,51200,0.6917840109931098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,4096,0.04897333184878031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,6144,0.04515377680460612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,4096,0.06707466973198785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,4096,0.03487911158137851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,3584,0.043143110142813794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,3072,0.05636355612013075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,3584,0.059466666645473905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,16384,0.11000355747010972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,3072,0.026416889495319788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,2560,0.0470142232047187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,3072,0.05287466777695549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,65536,0.9125715891520182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,3072,0.03792266713248359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,2048,0.038302222887674965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,2560,0.023281777898470562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,3584,0.030460443761613634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,2560,0.04626399940914578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,1536,0.029923554923799302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,2048,0.028149333265092637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,7168,0.11771110693613689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,1536,0.0206666671567493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,1024,0.021912889348136053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,1536,0.03352088729540507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,1536,0.01721155477894677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,1024,0.013791999883121915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,2560,0.03290044599109226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,768,0.01625688870747884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,1024,0.02703022294574314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,4096,0.0727528863483005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,768,0.013450666434235044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,512,0.011425777441925473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,768,0.025795555777019922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,2048,0.04005333450105455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,512,0.011361777782440186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,512,0.010699555277824402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,512,0.02260622216595544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,256,0.008642666869693333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,128,0.004252444538805219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,2048,0.019645333290100098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,128,0.007347555624114142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,256,0.007282666862010956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,128,0.009750222166379293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,64,0.003528888854715559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,32,0.0035866668654812705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,256,0.009721777505344814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,256,0.01646399994691213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,64,0.009111111362775167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,32,0.009340444372759925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,768,0.012186666329701742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,65536,0.3228444523281521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,65536,0.17950310972001818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,51200,0.2514382203420003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,3584,0.06413511435190837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,65536,0.1578577756881714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,51200,0.13154666953616673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,1024,0.01572177807490031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,51200,0.12178844875759548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,16384,0.04747555653254191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,16384,0.0420720014307234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,16384,0.05274933245446947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,128,0.013597333596812354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,65536,0.18136800660027397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,10240,0.0528675549560123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,16384,0.09152888589435154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,12288,0.035832888550228544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,51200,0.13573866420321995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,8192,0.04443822304407755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,12288,0.07181244426303439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,10240,0.03090488910675049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,8192,0.027958220905727808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,12288,0.03660088777542114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,7168,0.03895555602179633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,8192,0.03065955638885498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,12288,0.04025244381692674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,6144,0.03313333458370633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,7168,0.028047998746236164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,7168,0.03047377864519755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,10240,0.031756444109810725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,7168,0.02846577763557434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,10240,0.03523644473817613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,6144,0.02676444417900509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,4096,0.02403822210099962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,5120,0.017874666386180453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,5120,0.023510222633679707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,4096,0.020288000504175823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,4096,0.016176889340082806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,3584,0.02102577851878272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,3584,0.018283555905024212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,3584,0.019139554765489366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,3584,0.014946666028764514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,5120,0.024471999870406255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,8192,0.024870221813519795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,3072,0.016536000702116225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,3072,0.013457777599493662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,3072,0.01725955473052131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,5120,0.029020445214377508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,2560,0.01586666703224182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,2048,0.013190222283204397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,2560,0.012156444291273752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,2048,0.01313688854376475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,2560,0.0158942225906584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,4096,0.021341333786646526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,2048,0.012817777693271637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,6144,0.02656800051530202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,1536,0.011161777708265515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,1536,0.01313511116637124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,3072,0.0185199992524253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,1024,0.009397333694828881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,2048,0.015214223000738354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,1024,0.011535999675591787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,768,0.0070062221752272705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,1536,0.0107022225856781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,768,0.007753777835104201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,768,0.007607111500369177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,512,0.004848888764778773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,768,0.010103111465771994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,1024,0.008440000315507254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,1536,0.010737777584128909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,512,0.007355555891990662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,512,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,512,0.00867733359336853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,256,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,256,0.005850666513045629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,256,0.008041777544551426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,128,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,64,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,2560,0.015150222513410779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,32,0.0028453332682450614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,256,0.0032008888406885993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,64,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,32,0.0063279999627007385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,65536,0.2550862100389269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,1024,0.008352888955010308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,65536,0.15197866492801243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,128,0.007032000356250339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,51200,0.19891111056009927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,65536,0.13559199704064265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,16384,0.06746221913231744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,51200,0.11275733841790093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,51200,0.10311466455459595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,16384,0.04126755396525065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,128,0.003170666595300039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,12288,0.052778667873806424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,16384,0.04394666684998406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,65536,0.13968177636464438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,10240,0.04498222139146593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,51200,0.11004177729288737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,6144,0.02014400064945221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,10240,0.027886221806208294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,10240,0.030202666918436687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,8192,0.03804088963402642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,16384,0.037901331981023155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,8192,0.024352888266245525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,12288,0.029890666405359905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,128,0.007635555333561367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,7168,0.033125334315829806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,12288,0.03458311160405477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,8192,0.02674666709370083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,6144,0.029314666986465454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,7168,0.020065777831607394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,7168,0.02630222174856398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,7168,0.02492888934082455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,10240,0.02717422280046675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,6144,0.022508444057570562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,5120,0.021200888686709937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,12288,0.031996445523367986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,4096,0.022300443715519373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,5120,0.02033155494266086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,5120,0.01625777781009674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,4096,0.014494222071435718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,4096,0.018623110320832994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,3584,0.019845333364274766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,4096,0.01791733337773217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,6144,0.024363555841975745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,3072,0.017882666654056974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,3584,0.013428444663683573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,3584,0.016504888733228047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,3584,0.016768000192112394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,3072,0.012690666649076672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,3072,0.01570666664176517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,2560,0.013855111267831592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,2560,0.011382222175598145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,2560,0.014182221558358936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,5120,0.025754666990704004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,2048,0.012087110843923358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,1536,0.009729777773221334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,2048,0.013434666726324292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,2560,0.012598222328556908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,3072,0.015121777852376303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,1536,0.009759110709031424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,1536,0.012034667034943899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,2048,0.01074044406414032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,1024,0.007948444121413762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,1024,0.007991111112965478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,1024,0.010713777608341642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,2048,0.009924444059530893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,768,0.007264888948864407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,768,0.0073315559162033936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,768,0.009000889129108852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,1024,0.00776711106300354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,1536,0.010203555226325989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,256,0.0038817777401871155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,512,0.007296889192528195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,512,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,512,0.008672888908121321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,128,0.0032426667296224167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,768,0.006362666686375936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,256,0.007742221984598372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,128,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,128,0.007009777757856581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,64,0.0030862221287356485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,32,0.003212444484233856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,8192,0.021872000561820135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,64,0.006139555739031897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,32,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,256,0.006716444260544247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,6144,0.018218666315078735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,65536,0.1396204498079088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,51200,0.1742266681459215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,65536,0.12643110752105713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,512,0.0042257776690853965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,51200,0.1079182227452596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,16384,0.06036000119315254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,51200,0.09794311390982734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,65536,0.21871199872758654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,16384,0.039488001002205744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,65536,0.13218666447533503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,256,0.0059004442559348206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,16384,0.03619911273320516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,16384,0.04049866729312473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,51200,0.10551822185516357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,10240,0.038968000147077776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,12288,0.02882400155067444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,12288,0.03054844339688619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,12288,0.03272533416748047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,10240,0.026734222968419392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,10240,0.02864977717399597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,7168,0.029615110821194116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,8192,0.02309600015481313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,8192,0.02539377742343479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,12288,0.04715111189418369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,6144,0.02645422187116411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,8192,0.0328400002585517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,128,0.007669332954618666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,6144,0.020629333125220407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,5120,0.022967999180157978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,6144,0.01732711162832048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,8192,0.021360889077186584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,6144,0.020980444219377305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,5120,0.01609866652223799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,7168,0.02260355485810174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,4096,0.01980533368057675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,7168,0.02325511144267188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,4096,0.01443733274936676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,3584,0.017717333303557504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,3584,0.01477599971824222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,5120,0.018295110927687753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,5120,0.019899555378490023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,7168,0.01966399947802226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,3584,0.01628266606065962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,4096,0.01518133282661438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,10240,0.025985777378082275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,3072,0.013089777694808112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,2560,0.012461333639091916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,3072,0.011955555942323474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,2560,0.010592888626787396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,2560,0.011952000359694162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,2560,0.013793778088357715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,2048,0.009350222018029954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,2048,0.010656000011497073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,2048,0.013398222625255585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,4096,0.017080000705189176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,1536,0.009359999663299983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,1024,0.007622222105662028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,1536,0.01201600001917945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,2048,0.011061333119869232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,1024,0.007423111134105259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,1024,0.007355555891990662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,1024,0.010085333552625444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,3584,0.012996444271670448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,768,0.0069848886794514125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,512,0.0047866664826869965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,768,0.009019555317031013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,1536,0.00867377801073922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,512,0.006039111150635614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,512,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,256,0.003896000070704354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,512,0.00835111074977451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,768,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,128,0.0031991110493739447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,256,0.006576889091067844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,3072,0.016006223029560514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,256,0.007420444654093848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,64,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,128,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,32,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,3072,0.015247111519177755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,64,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,65536,0.1749217775132921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,32,0.005896889087226655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,65536,0.08950755331251357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,256,0.005585777676767773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,1536,0.009016888837019602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,65536,0.07271999782986112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,768,0.007344000041484833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,128,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,16384,0.04595822095870972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,51200,0.0696657763587104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,51200,0.05810844235950046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,12288,0.03295022249221802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,65536,0.13240977128346762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,16384,0.027489778068330552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,16384,0.04045244389110141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,12288,0.021653332644038733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,12288,0.032249778509140015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,51200,0.13129333655039468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,10240,0.018927999668651156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,8192,0.023637332850032385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,10240,0.017495110630989075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,128,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,16384,0.023740443918440077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,8192,0.016537777251667447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,7168,0.02055999967787001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,8192,0.024561777710914612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,51200,0.1054311129781935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,12288,0.019179556104871962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,7168,0.013792888985739814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,7168,0.017782222893502977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,7168,0.022408000297016565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,5120,0.015863110621770222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,6144,0.015818667080667283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,6144,0.020427554845809937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,10240,0.028622223271263972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,5120,0.014631999863518609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,4096,0.013596444494194455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,5120,0.018394667241308425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,8192,0.014888889259762235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,3584,0.012093333734406365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,4096,0.010975999964608086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,10240,0.02777333392037286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,4096,0.016767111089494493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,3072,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,6144,0.012771555946932899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,3584,0.015480000111791821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,3072,0.011033777561452655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,2560,0.009775111244784461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,4096,0.013097777962684631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,3584,0.009962666365835402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,2560,0.013765333427323235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,2560,0.008535111116038429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,2048,0.007967111137178209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,6144,0.01828888886504703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,2048,0.009289777941173976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,2048,0.007967111137178209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,1536,0.007636444436179266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,2048,0.012991999586423239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,5120,0.011737777955002256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,1536,0.007535999847782983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,1024,0.006038222048017714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,1536,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,2560,0.010346666806273991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,3584,0.012164444559150271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,1024,0.00674311113026407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,768,0.004596444467703502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,1024,0.009352000223265754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,768,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,768,0.005995555470387141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,512,0.0038862224254343244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,768,0.008719999757077958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,3072,0.014461333552996317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,256,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,1536,0.01071911139620675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,512,0.005924444645643234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,512,0.008039111064540016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,256,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,256,0.0053013331360287136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,1024,0.006607111129495833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,64,0.002891555635465516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,128,0.005986666513813867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,32,0.0028568889117903183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,128,0.007548444800906711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,64,0.005583999885453119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,65536,0.1496986680560642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,3072,0.009104000197516548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,32,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,512,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,65536,0.08282133605745104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,65536,0.06671377685334948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,256,0.007648000286685095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,51200,0.06323644187715319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,51200,0.05403377612431844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,65536,0.13208799892001682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,51200,0.1051902241177029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,16384,0.022034666604465906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,51200,0.11755643950568305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,128,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,12288,0.019878221882714164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,10240,0.028582221931881372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,16384,0.0421377784676022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,12288,0.03243111239539252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,10240,0.017253332667880587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,10240,0.01622577839427524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,16384,0.026050666968027752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,10240,0.028022223048739966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,12288,0.03337600164943271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,7168,0.018930666976504855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,8192,0.024570667081409033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,12288,0.01827555563714769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,7168,0.017606221967273288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,128,0.004959110998445087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,7168,0.013429332938459186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,7168,0.02232444451914893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,8192,0.021958221991856892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,6144,0.01514488955338796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,5120,0.014497778481907315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,6144,0.01219377749496036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,16384,0.040056887600157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,8192,0.015114666687117683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,5120,0.013677332964208392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,5120,0.018414222531848483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,4096,0.009703111317422655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,6144,0.016527111331621807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,3584,0.011028444601429833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,4096,0.01622577839427524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,3584,0.009363555245929295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,3584,0.011401777466138205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,3584,0.015504888362354703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,6144,0.02030311193731096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,4096,0.01292711082432005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,3072,0.00853777759604984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,3072,0.014480888843536377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,4096,0.012088889049159156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,2560,0.009472000102202097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,2048,0.007671111159854465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,2560,0.007717333734035492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,2560,0.013334222137928009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,3072,0.010301333334710863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,2048,0.007594666547245449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,1536,0.0071937781241205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,2048,0.01260088880856832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,8192,0.014142221874660917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,3072,0.01072266697883606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,1024,0.0052391112678580815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,1536,0.007183999651008182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,1536,0.01036533299419615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,768,0.004576000074545543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,1024,0.006314666734801398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,1024,0.009009777671760982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,768,0.005972444597217772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,768,0.008788444101810455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,512,0.0035457776652442086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,5120,0.011161777708265515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,2560,0.009038222332795462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,1536,0.00703288863102595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,256,0.0032551110618644287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,512,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,1024,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,512,0.008026666939258575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,256,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,256,0.005306666509972678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,256,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,2048,0.008408889174461365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,128,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,64,0.0028986665937635633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,128,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,64,0.0053084443012873335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,32,0.005315555466545953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,65536,0.12325955761803521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,128,0.0028693332440323303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,65536,0.07575377490785387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,128,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,65536,0.06428177489174737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,32,0.0027217778066794076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,16384,0.03621955712636312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,65536,0.1318942175971137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,512,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,16384,0.02148088812828064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,16384,0.023641778363121882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,51200,0.10460088650385539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,768,0.0062977779242727495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,12288,0.018343110879262287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,51200,0.09801244735717773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,10240,0.02502577834659153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,51200,0.051779554949866406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,51200,0.0567528870370653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,8192,0.021751999855041504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,10240,0.027828445037206013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,16384,0.03952444593111674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,12288,0.028823110792371962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,8192,0.014128888646761576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,7168,0.019113777412308585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,8192,0.013922666509946188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,8192,0.02402399977048238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,6144,0.01724799970785777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,7168,0.012882666455374824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,12288,0.017437333861986797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,7168,0.02202577723397149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,5120,0.01538133290078905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,6144,0.011946666571829053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,6144,0.014494222071435718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,6144,0.019908444748984445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,10240,0.015673778123325773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,5120,0.013171555267439948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,5120,0.017877333694034152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,4096,0.009690666364298927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,3584,0.010035555395815108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,7168,0.016787555482652452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,3584,0.01073066641887029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,3584,0.015257777439223396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,4096,0.011720000041855706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,3072,0.010144000252087912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,4096,0.01628444426589542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,12288,0.03166666626930237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,3072,0.008351999852392409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,3072,0.014142221874660917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,3584,0.009020444419648912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,5120,0.010646222366227044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,2048,0.007445333732499017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,2560,0.007641777396202087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,4096,0.011734222372372946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,2048,0.007707555260923173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,1536,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,2048,0.011765333513418833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,2560,0.008622222476535374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,1536,0.00664533343580034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,1536,0.010207110808955299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,10240,0.015973332855436537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,2560,0.013438222308953604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,1024,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,768,0.004332444320122401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,2048,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,768,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,1536,0.006989333364698622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,768,0.006264888991912206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,512,0.003671110918124517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,3072,0.009233777721722921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,512,0.006225777582989798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,512,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,256,0.0032497776879204642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,512,0.008044444852405125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,1024,0.009416888985368941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,128,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,256,0.005990222096443176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,256,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,256,0.007668444679843054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,128,0.005024888863166173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,128,0.00757155567407608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,1024,0.004757333132955763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,32,0.0028364445186323593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,64,0.005471999860472149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,1024,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,65536,0.10747288995318943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,65536,0.06772977775997586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,128,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,64,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,51200,0.08580888642205133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,65536,0.061342219511667885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,65536,0.1316479974322849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,16384,0.03240622083346049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,51200,0.04971288972430759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,32,0.005296000176005893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,51200,0.10507733292049831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,16384,0.021220443977249995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,16384,0.03917600048912896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,12288,0.01657777859105004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,12288,0.03125600020090739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,10240,0.022435555855433147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,12288,0.016880000631014507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,10240,0.015056888262430826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,10240,0.01495822270711263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,12288,0.026147555973794725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,8192,0.013722666435771518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,2560,0.008986666798591614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,8192,0.013238222234778933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,7168,0.01755466726091173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,51200,0.05148355497254265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,7168,0.01661422186427646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,7168,0.012424888710180918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,6144,0.015858666764365304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,8192,0.019853333632151287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,16384,0.02058577703105079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,5120,0.014186667071448432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,6144,0.011400000088744693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,6144,0.01367555558681488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,6144,0.019831111033757527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,4096,0.010829333629873065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,5120,0.009920000202125972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,5120,0.01825777855184343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,4096,0.011558222273985544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,3584,0.009690666364298927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,4096,0.016172443826993305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,7168,0.021792888641357422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,10240,0.02731200059254964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,3072,0.008655110994974772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,3584,0.010413332945770687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,3584,0.015033778217103748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,768,0.008805333740181392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,2560,0.0076657773719893555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,4096,0.00906844437122345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,2560,0.008663111262851292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,2560,0.007672000262472365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,2560,0.013436444103717804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,8192,0.023804444405767653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,2048,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,3072,0.014160888062583076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,1536,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,2048,0.007071110937330458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,2048,0.011534222298198275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,1536,0.0068604445291890045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,1536,0.006816888848940532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,3584,0.008656000097592672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,1536,0.010400888820489248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,1024,0.006353777729802662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,768,0.004274666723277834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,5120,0.012448888685968188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,1024,0.009100444614887238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,768,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,512,0.0035342222286595237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,768,0.00866844422287411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,1024,0.004586666822433472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,512,0.005611555443869696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,256,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,512,0.005968889014588461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,1024,0.006219555520349079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,3072,0.009647111097971598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,256,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,128,0.0028728888266616394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,256,0.007384888827800751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,768,0.005633777628342311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,128,0.0059164443777667145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,128,0.004936000125275718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,32,0.0028355556229750314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,64,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,512,0.008368888662921058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,65536,0.0864586697684394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,3072,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,256,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,65536,0.061307556099361844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,51200,0.06780266761779785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,65536,0.05922933419545492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,64,0.0026773332307736077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,65536,0.13120710849761963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,16384,0.026369776990678575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,51200,0.04721866713629829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,2048,0.0070186663005087115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,16384,0.019620445039537217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,12288,0.02033955521053738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,16384,0.039537777503331505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,12288,0.015463999576038785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,12288,0.031559109687805176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,10240,0.018411555223994784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,128,0.007271111011505127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,51200,0.04790666699409485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,10240,0.013785777820481194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,8192,0.016176000237464905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,51200,0.10493955347273086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,16384,0.019144889381196763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,10240,0.02734222180313534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,7168,0.014152889450391134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,8192,0.012695999609099494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,8192,0.012415111064910889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,8192,0.02358755634890662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,6144,0.013001778059535556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,7168,0.011759999725553723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,12288,0.01620533400111728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,7168,0.021564443906148274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,5120,0.011922666596041786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,6144,0.010739555789364709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,6144,0.012356444365448423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,6144,0.019955555597941082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,4096,0.012427555190192329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,32,0.005229333208666907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,5120,0.017924444542990792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,4096,0.010657777388890585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,3584,0.011422221859296164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,4096,0.0158924443854226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,7168,0.014496889379289416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,3584,0.00811644477976693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,3584,0.015071110592948066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,5120,0.009702222214804756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,5120,0.011395555403497485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,3072,0.008181333541870117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,3072,0.007843555675612556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,3072,0.013960000541475085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,4096,0.008720888859695857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,2560,0.007481777833567725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,2560,0.007896888587209914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,2560,0.012069332930776807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,2048,0.007071110937330458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,1536,0.00489155575633049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,2048,0.01108444482088089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,10240,0.014577777849303352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,2560,0.007610666255156199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,1024,0.004224888980388641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,2048,0.00638133328821924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,1024,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,2048,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,3072,0.01036088913679123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,1024,0.006275555739800136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,768,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,768,0.0087013335691558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,1536,0.00665511108107037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,512,0.0032319999817344877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,768,0.006189333481921091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,512,0.005962666538026597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,512,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,512,0.008016889293988546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,3584,0.009346666435400644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,768,0.003551111039188173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,1024,0.009324444664849175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,128,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,256,0.005962666538026597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,256,0.007672888537247975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,128,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,64,0.002571555475393931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,32,0.002817777709828483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,1536,0.0063324446479479475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,65536,0.07873333162731595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,64,0.005220444252093633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,32,0.005227555417352253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,256,0.003198222153716617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,51200,0.06091555621888903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,65536,0.058435552650027804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,65536,0.05815822548336453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,256,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,65536,0.13144800398084852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,16384,0.02369955513212416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,51200,0.04548888736300998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,128,0.004951999833186467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,51200,0.10465333196851943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,1536,0.010111110905806223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,16384,0.01833599971400367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,16384,0.03915466533766852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,10240,0.016865778300497267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,12288,0.014858666393491956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,12288,0.031211555004119873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,10240,0.013379555609491138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,10240,0.027316444449954565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,51200,0.046907554070154824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,12288,0.01959822244114346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,16384,0.019354666272799175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,7168,0.013455111119482251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,8192,0.012420444852775998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,8192,0.02358666724628872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,12288,0.016136889656384785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,6144,0.01276000009642707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,7168,0.01107911103301578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,7168,0.013462222284740873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,7168,0.02167911165290409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,10240,0.014472888575659858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,8192,0.014673777752452426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,5120,0.011738667057620155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,6144,0.019596444235907663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,5120,0.009368889033794403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,128,0.00741511086622874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,4096,0.011403555671374003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,5120,0.017887110511461895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,4096,0.00851644410027398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,3584,0.01111644423670239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,6144,0.012102222277058495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,6144,0.010482666393121084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,3584,0.00870400004916721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,3072,0.010037333601050908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,8192,0.012396444049146442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,3584,0.01498044364982181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,3072,0.008029333419269985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,3072,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,2560,0.00872444444232517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,3072,0.014119999276267158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,4096,0.015488889482286243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,2048,0.005607999861240387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,2560,0.007345777418878343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,2560,0.0074924445814556545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,2560,0.012479999827014076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,1536,0.004609777695602841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,2048,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,2048,0.006509333435032103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,3584,0.008027555214034187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,1024,0.003934222377008862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,1536,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,1536,0.006703111032644908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,1536,0.010037333601050908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,1024,0.00630044440428416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,1024,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,768,0.005608888963858287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,768,0.005991111199061076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,768,0.008340444829728868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,512,0.0052800000541739995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,512,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,256,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,768,0.0035777779089079965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,1024,0.0058062221441004015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,256,0.005919111271699269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,128,0.0028968888024489083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,256,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,128,0.004973333328962326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,2048,0.011071110765139261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,64,0.0028488888508743713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,128,0.005920889063013925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,512,0.005968889014588461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,4096,0.009371555513805812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,64,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,65536,0.06320622232225206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,32,0.004949333353175057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,512,0.0033982222278912864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,51200,0.05086311035686069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,65536,0.13145866658952501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,5120,0.010712888505723743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,128,0.007679999702506595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,51200,0.04351377818319532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,16384,0.02067733307679494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,51200,0.04602666695912679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,51200,0.10490222109688653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,16384,0.01754666699303521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,12288,0.01757244434621599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,16384,0.03921866748068068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,65536,0.05586577786339653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,12288,0.014465777410401238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,10240,0.015208888385030957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,12288,0.03109955456521776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,10240,0.01384622189733717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,10240,0.013108444710572561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,10240,0.027228444814682007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,16384,0.019125334090656705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,8192,0.011757333245542316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,7168,0.012446222205956778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,8192,0.011965333587593503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,65536,0.05726044707828098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,7168,0.012092444631788464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,6144,0.011462222370836469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,7168,0.02127644419670105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,32,0.0028408887899584244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,6144,0.011284444067213269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,5120,0.010492444038391113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,256,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,8192,0.023374223046832617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,5120,0.009666666388511658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,5120,0.009407999614874521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,5120,0.01721600029203627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,7168,0.0107342220014996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,4096,0.00812177773978975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,4096,0.008370666868156856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,3584,0.009713778065310584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,4096,0.0151555554734336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,6144,0.010229333407349056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,3072,0.009300444689061906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,3584,0.008034666379292807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,3584,0.014112000664075216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,12288,0.015844444433848064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,4096,0.009718221922715506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,3072,0.01277955538696713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,3072,0.007716444631417592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,8192,0.013077333569526672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,2560,0.006781333436568578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,2048,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,2560,0.007602666815121968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,2560,0.012431999875439538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,6144,0.01959111127588484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,3584,0.007686222592989604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,2048,0.007027555671003129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,2048,0.006351111249791251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,2048,0.01073066641887029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,3072,0.0073520003093613525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,1536,0.006802666518423293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,2560,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,1536,0.010044444766309526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,768,0.003567111161020067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,1024,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,1024,0.008807111117574904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,768,0.005468444277842839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,512,0.0032293332947625052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,1024,0.0038684445122877755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,1536,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,512,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,1024,0.00565422202150027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,512,0.008289777570300633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,768,0.006014222072230444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,768,0.00868266655339135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,256,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,128,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,128,0.00741511086622874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,128,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,256,0.003007111036115222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,32,0.0028106667515304354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,64,0.005236444373925527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,256,0.005616888817813661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,1536,0.00489777781897121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,65536,0.05306222372584873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,65536,0.1308737728330824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,64,0.0026497778793176017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,51200,0.04565688967704773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,51200,0.04160355528195699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,32,0.00509688879052798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,16384,0.018996444013383653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,51200,0.1042133304807875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,51200,0.04526844289567736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,512,0.0053084443012873335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,16384,0.018998222218619455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,12288,0.016350222958458793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,65536,0.05607555309931437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,12288,0.014137778017255994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,12288,0.01517599986659156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,10240,0.013762666947311826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,12288,0.03117155697610643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,256,0.005133333305517833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,8192,0.01254488858911726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,10240,0.013474666410022311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,10240,0.01277511070171992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,10240,0.027099554737408955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,7168,0.011697777443461947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,8192,0.01145155562294854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,16384,0.01720888912677765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,8192,0.02369244396686554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,6144,0.010188444621033138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,16384,0.03865155577659607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,7168,0.010374222364690568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,7168,0.021339555581410725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,5120,0.00978133330742518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,6144,0.010851555400424533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,6144,0.019501333435376484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,128,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,4096,0.009081777599122789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,5120,0.017147555947303772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,7168,0.011295999917719098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,4096,0.0075795559419526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,3584,0.009009777671760982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,6144,0.010072000324726105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,4096,0.008378666308191087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,65536,0.05839911434385511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,3584,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,3584,0.014135110709402295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,8192,0.011095999843544431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,3072,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,2560,0.005917333480384614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,3072,0.013090666797426013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,4096,0.015208888385030957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,2560,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,2048,0.004955555415815777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,2560,0.0075031113293435835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,2560,0.011753777662913004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,3584,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,1536,0.004576888763242298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,2048,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,2048,0.006264888991912206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,5120,0.009357333183288574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,2048,0.011120888921949597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,1536,0.006676444576846228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,3072,0.007663999994595845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,1536,0.009866666462686326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,1024,0.00565422202150027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,1024,0.0063928887248039246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,768,0.003565333369705412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,3072,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,768,0.00886222223440806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,768,0.006014222072230444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,5120,0.009684444301658207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,1024,0.0038844446341196695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,1536,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,256,0.0031964443624019623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,512,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,512,0.008393777741326226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,1024,0.00903555585278405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,128,0.0028657778683635923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,256,0.004943999979231092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,256,0.007703111403518253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,768,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,64,0.002509333400262727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,128,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,512,0.0032355555643637977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,128,0.007305777735180325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,65536,0.05081955591837565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,64,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,512,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,32,0.005296000176005893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,51200,0.03707199957635667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,65536,0.055494222376081676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,256,0.005868444426192178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,65536,0.13084267245398626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,16384,0.015963556038008798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,51200,0.040580444865756564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,51200,0.1042906641960144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,128,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,12288,0.013080000049538083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,16384,0.0165955556763543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,16384,0.03867200016975403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,10240,0.012129777835475074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,12288,0.01425333321094513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,12288,0.013916444447305469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,12288,0.030631999174753826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,8192,0.010715555813577441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,10240,0.012448000411192576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,10240,0.02677511175473531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,51200,0.04473511046833462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,7168,0.01071111112833023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,8192,0.010727110836240979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,8192,0.023431999815834895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,16384,0.01864266726705763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,32,0.00254488881263468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,7168,0.00979644474056032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,7168,0.021249777740902368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,6144,0.009022222624884712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,6144,0.009920000202125972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,6144,0.018681777848137748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,5120,0.008354666332403818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,5120,0.01685422162214915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,4096,0.010067555639478896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,5120,0.009342222577995723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,6144,0.012597333225939008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,7168,0.010414222048388587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,3584,0.00962755580743154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,4096,0.008310221963458592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,4096,0.007344000041484833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,65536,0.050810668203565806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,5120,0.011481777661376528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,3584,0.008036444584528605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,3584,0.007348444726732042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,3584,0.013990221752060784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,2560,0.005620444400442972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,3072,0.006919999917348226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,3072,0.01312888910373052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,2560,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,2048,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,2560,0.012105777859687805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,4096,0.01475911173555586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,10240,0.01311555587583118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,1536,0.004557333472702238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,2048,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,3072,0.007693332930405934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,8192,0.011061333119869232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,1024,0.003914666672547658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,1536,0.006890666567616993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,1536,0.010049777726332346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,2560,0.006647999915811751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,768,0.003552000141806073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,1024,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,1024,0.009381333159075843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,3072,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,768,0.005352889084153705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,2048,0.01109955542617374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,1536,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,512,0.005374222166008419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,512,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,512,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,1024,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,256,0.005028444445795483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,256,0.005897777775923411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,128,0.0028702221396896574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,256,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,768,0.006211555666393704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,512,0.00328533330725299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,768,0.008361777497662438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,128,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,128,0.004951111144489712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,32,0.002903999967707528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,128,0.007328000333574083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,64,0.004947555561860402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,256,0.0028942223224374983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,32,0.004940444396601783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,51200,0.02938400043381585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,65536,0.05511200096872118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,65536,0.13089244895511204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,2048,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,51200,0.03904089000489976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,64,0.002593777866827117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,51200,0.04446844591034783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,51200,0.104476445251041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,12288,0.01166488892502255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,16384,0.0162808895111084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,16384,0.038423998488320246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,65536,0.04900622367858887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,12288,0.01346577786737018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,10240,0.01071111112833023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,16384,0.013271999855836233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,10240,0.012081777883900536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,8192,0.009560888840092553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,10240,0.027106665902667578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,10240,0.012793777717484368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,16384,0.01704266667366028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,7168,0.0099653336736891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,8192,0.011092444260915121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,8192,0.010026666853162978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,8192,0.022673777408070032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,12288,0.013460444079505073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,7168,0.009406222237481011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,12288,0.03086400032043457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,7168,0.020593777298927307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,5120,0.00796888851457172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,6144,0.009692444569534725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,6144,0.008941333327028487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,4096,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,5120,0.009024000002278222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,5120,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,5120,0.016367110941145156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,65536,0.03632622294955783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,4096,0.008309333688682979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,4096,0.007409777906205919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,4096,0.014151111245155334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,3584,0.008017777568764156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,3584,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,7168,0.010096888575288985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,3584,0.013839999834696451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,2560,0.005774222314357758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,3072,0.007796444826655918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,3072,0.006980444822046492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,3072,0.013146667016877068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,2560,0.007382222347789341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,2560,0.011563555234008364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,3584,0.007384888827800751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,2048,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,2048,0.0070088886552386815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,3072,0.00657155571712388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,6144,0.008971555365456475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,1536,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,1024,0.003896000070704354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,1536,0.009704888694816166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,6144,0.018851555056042142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,2048,0.006322666588756774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,768,0.003555555724435382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,1024,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,1024,0.00869955536392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,1536,0.004794666750563515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,768,0.00601333338353369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,768,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,1536,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,512,0.004982222285535601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,256,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,512,0.005913777897755305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,2560,0.00646666685740153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,1024,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,256,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,128,0.0028560000161329904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,256,0.007634667058785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,768,0.00563733321097162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,512,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,128,0.005663110978073544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,2048,0.01071999967098236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,32,0.002534222271707323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,65536,0.025828444295459326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,64,0.004933333231343163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,32,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,512,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,51200,0.02160888910293579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,65536,0.046967999802695386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,65536,0.05464533302519056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,256,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,51200,0.04417511158519321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,51200,0.03774844606717428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,64,0.002551111082235972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,51200,0.1039502223332723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,128,0.004961777892377642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,12288,0.012424888710180918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,16384,0.015885333220163982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,16384,0.038346668084462486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,12288,0.013428444663683573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,12288,0.01311466677321328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,10240,0.011222222612963783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,12288,0.030511998467975195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,65536,0.13128177324930826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,16384,0.014827556080288358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,10240,0.01110577748881446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,10240,0.026012443833880957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,8192,0.010054222411579555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,7168,0.009725333087974125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,7168,0.009377777576446533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,8192,0.010602666272057427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,7168,0.020297777321603563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,6144,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,16384,0.0151591118839052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,6144,0.00868800034125646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,6144,0.018234666850831773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,10240,0.012070222033394707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,5120,0.007991111112965478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,7168,0.01036000003417333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,5120,0.008661333057615492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,4096,0.007636444436179266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,8192,0.011058666639857821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,4096,0.007913777397738563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,4096,0.008365333080291748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,4096,0.014479110638300577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,6144,0.009395555489593083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,3584,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,3584,0.013839110732078552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,3584,0.007654222349325816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,5120,0.008309333688682979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,128,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,2560,0.005920889063013925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,3072,0.012427555190192329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,3584,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,2560,0.006692444284756978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,2048,0.006696888970004187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,2560,0.011409777734014722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,3072,0.0063235556913746735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,2048,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,2048,0.010592000352011787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,3072,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,3072,0.007693332930405934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,1536,0.0068400001360310455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,1536,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,1024,0.004524444540341695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,1536,0.009823111196359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,2560,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,768,0.0042239998777707415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,1024,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,1024,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,1024,0.009156444834338294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,512,0.003588444242874781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,2048,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,1536,0.005655999812814925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,512,0.005305777821275923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,256,0.0032266666077905228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,512,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,256,0.004957333207130432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,256,0.00776711106300354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,128,0.0028515555378463534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,256,0.005658666706747479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,768,0.005281777845488654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,5120,0.016547555724779766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,128,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,64,0.0028435554769304064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,128,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,32,0.002862222285734283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,512,0.005991111199061076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,768,65536,0.022072000635994807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,64,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,32,0.004908444566859139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,768,51200,0.018315555320845712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,65536,0.0544453329510159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,65536,0.13059733973609075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,768,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,51200,0.037348445918824934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,768,16384,0.010658666491508484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,8192,0.022682666778564453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,51200,0.10338133573532104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,768,12288,0.010786666638321347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,16384,0.01616533266173469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,16384,0.014550222290886773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,16384,0.0373582210805681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,768,10240,0.009708444277445475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,12288,0.012086222569147745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,65536,0.04665155543221367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,12288,0.02980977627966139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,768,8192,0.009338666995366415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,10240,0.012101333174440594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,10240,0.025736888249715168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,768,0.008658666577604082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,8192,0.009734222458468543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,768,7168,0.008656000097592672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,8192,0.02237777743074629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,7168,0.009379555781682333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,12288,0.013443555268976422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,7168,0.020261334048377145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,768,6144,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,7168,0.010318222145239512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,6144,0.009692444569534725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,6144,0.00870222184393141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,51200,0.04356355468432108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,6144,0.018584888842370775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,8192,0.010724444356229572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,5120,0.009008000294367472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,768,4096,0.007648888561460707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,5120,0.01681066718366411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,4096,0.007384888827800751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,768,3584,0.007376000285148621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,10240,0.011076444553004371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,768,5120,0.007648888561460707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,3584,0.007324444750944774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,768,3072,0.007015110717879401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,3584,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,3584,0.013456888496875763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,3072,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,3072,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,3072,0.01276088919904497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,4096,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,4096,0.014177777700954013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,768,2048,0.005241777747869492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,2560,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,2560,0.011743110915025076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,768,1536,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,2048,0.006039111150635614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,2048,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,2048,0.011083555718262991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,768,1024,0.004247111164861255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,1536,0.00666222224632899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,1536,0.006042666733264923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,1536,0.009696889254781935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,768,768,0.0038764443662431505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,1024,0.0063271112740039825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,1024,0.009054222040706212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,2560,0.006637333167923822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,768,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,768,0.00871377769443724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,512,0.005281777845488654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,5120,0.008285333712895712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,768,256,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,512,0.008233778178691864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,768,2560,0.0064604443808396654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,1024,0.005642666584915585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,256,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,256,0.00776088900036282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,768,512,0.0032177778581778207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,128,0.004931555440028508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,128,0.005590222362014983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,128,0.007679999702506595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,128,0.0076657773719893555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,768,32,0.0026764445420768526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,512,65536,0.01793511046303643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,512,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,32,0.004954666727119022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,256,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,512,51200,0.015184889237085978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,65536,0.04587199952867296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,65536,0.05217688944604662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,768,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,512,16384,0.010886222124099731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,51200,0.03851022322972616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,768,64,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,51200,0.03711288836267259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,512,12288,0.008772444393899705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,51200,0.10382666852739121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,16384,0.015849777393870883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,64,0.004943999979231092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,16384,0.03740888833999634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,512,10240,0.008003555238246918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,12288,0.012309333516491784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,12288,0.029899554120169744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,10240,0.012135110795497894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,10240,0.011081777513027191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,65536,0.13037066989474827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,8192,0.009735110733244155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,512,7168,0.008011555506123437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,8192,0.022328888376553852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,16384,0.014509333504570855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,7168,0.009355555805895064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,512,6144,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,768,128,0.0028320000403457214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,512,8192,0.008107555409272512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,6144,0.008723555339707268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,6144,0.009701333112186855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,6144,0.018588445252842374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,5120,0.008367111285527548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,5120,0.016532444291644625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,5120,0.00905777762333552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,512,4096,0.007334222396214803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,4096,0.007698666718271043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,512,5120,0.007239999870459239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,512,3584,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,8192,0.01089511149459415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,3584,0.007328888608349695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,12288,0.013448889056841532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,7168,0.020301333732075162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,512,3072,0.006656000183688269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,3584,0.008000888758235509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,10240,0.02564711040920681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,4096,0.014847111370828418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,512,2560,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,3072,0.0074346669846110875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,2560,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,2560,0.011389333340856763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,512,2048,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,2560,0.007282666862010956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,2048,0.006064888917737537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,3072,0.007000000112586551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,512,1536,0.004545777622196409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,3072,0.0127697777416971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,1536,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,512,1024,0.003913777983850903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,1024,0.005617777920431561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,2048,0.010953777366214328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,4096,0.008383110993438298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,512,768,0.003637333297067218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,1024,0.009226666556464301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,7168,0.010040000081062317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,768,0.0053004444473319584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,768,0.008359111017651027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,768,0.005971555494599872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,512,0.005306666509972678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,512,0.008010666403505538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,512,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,1024,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,256,0.004951111144489712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,256,0.007616000043021307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,512,512,0.003228444399105178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,256,0.005647111270162795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,128,0.004959110998445087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,128,0.0074942219588491656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,2048,0.00692622239391009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,512,64,0.0028364445186323593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,128,0.005642666584915585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,3584,0.013821333646774292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,64,0.004581333448489507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,256,65536,0.015646222564909194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,512,128,0.0029102222373088202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,65536,0.04670133193333944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,65536,0.04411822226312426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,256,51200,0.01350844403107961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,65536,0.12909777959187826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,512,256,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,512,32,0.0028888889484935333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,51200,0.03461422191725837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,51200,0.1023591094546848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,1536,0.010061333576838175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,16384,0.014510222607188754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,256,12288,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,16384,0.03668266534805298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,12288,0.013188444077968597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,256,10240,0.007670222057236566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,256,16384,0.008683555656009251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,51200,0.03746844331423441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,10240,0.012099555797047086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,1536,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,256,8192,0.007275555696752336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,16384,0.015862221519152325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,8192,0.010861333045694562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,8192,0.009697777529557547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,256,7168,0.006672888994216919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,8192,0.02200088898340861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,12288,0.029517332712809246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,256,6144,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,7168,0.010328888893127441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,7168,0.009392889009581672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,7168,0.020288000504175823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,256,5120,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,6144,0.008676444490750631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,6144,0.018227555685573153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,10240,0.02551288902759552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,256,4096,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,5120,0.009003555609120263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,5120,0.016524444023768108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,12288,0.012208888928095499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,256,3584,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,4096,0.014483556151390076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,10240,0.010739555789364709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,3584,0.007329777710967594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,256,3072,0.006031111296680238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,3584,0.007669332954618666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,32,0.004975999808973736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,3584,0.013783999615245394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,256,2560,0.006440889090299606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,3072,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,3072,0.012110222544935016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,4096,0.00831911133395301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,2560,0.0069919998447100324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,256,2048,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,2560,0.012215110990736218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,6144,0.009357333183288574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,2048,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,256,1536,0.004563555535342959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,2048,0.010817777779367236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,5120,0.008178667061858708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,3072,0.007009777757856581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,1536,0.009721777505344814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,1536,0.006647999915811751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,4096,0.007034666836261749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,1024,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,1024,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,1024,0.009010666774378883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,2048,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,768,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,256,512,0.0032293332947625052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,1536,0.005991999887757831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,256,1024,0.003914666672547658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,512,0.005148444324731827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,256,256,0.0028737777223189673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,512,0.00831733312871721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,256,768,0.003535110917356279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,256,0.005617777920431561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,256,128,0.002862222285734283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,768,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,768,0.008657777474986183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,128,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,128,0.005759999983840519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,256,64,0.002577777744995223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,256,32,0.002536888958679305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,512,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,128,65536,0.014846222268210517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,64,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,32,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,128,51200,0.01275111072593265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,65536,0.04612266686227587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,65536,0.12844710879855686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,51200,0.034103999535242714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,128,16384,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,51200,0.10145422485139634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,128,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,2560,0.006334222025341458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,128,12288,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,16384,0.015861334072219003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,16384,0.03725333346260919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,256,0.0047582222355736625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,12288,0.012137778103351593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,128,10240,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,12288,0.01332622187005149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,256,0.007630222373538547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,128,8192,0.0070266665683852295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,10240,0.011917333636018964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,10240,0.011060444845093621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,10240,0.025430222352345783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,128,7168,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,8192,0.00982844415638182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,8192,0.022341334157519873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,16384,0.014483556151390076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,128,6144,0.006258666515350342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,7168,0.009121778110663096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,7168,0.010043555663691627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,7168,0.02032888929049174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,128,5120,0.006305777778228124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,6144,0.008763555851247575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,6144,0.018242667118708294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,51200,0.03721955418586731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,128,4096,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,5120,0.008645333349704742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,8192,0.010704889065689512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,5120,0.016520889268981088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,128,3584,0.005997333261701796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,4096,0.008160889148712158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,4096,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,4096,0.014315555493036905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,128,3072,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,3584,0.007300444775157505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,6144,0.009545777406957414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,12288,0.029159112109078303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,128,2560,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,3072,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,3072,0.012434666355450949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,5120,0.0081386665503184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,2560,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,2560,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,2560,0.011797332929240333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,2048,0.0069875551594628235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,3584,0.007975111405054728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,2048,0.011077333655622272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,3584,0.013799111048380533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,1536,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,128,1024,0.0038746665749284956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,128,2048,0.006039999839332368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,1024,0.005583999885453119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,1024,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,128,768,0.0037164443896876443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,65536,0.04252622193760342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,1536,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,1536,0.00982488857375251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,128,512,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,768,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,512,0.005247999810510212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,512,0.008360889222886827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,128,256,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,3072,0.006731555693679386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,768,0.008373333348168267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,256,0.007603555917739868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,256,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,128,128,0.0028568889117903183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,768,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,128,0.004582222137186262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,128,0.005623999983072281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,128,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,128,64,0.002518222149875429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,128,32,0.002560888934466574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,2048,0.006021333237489064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,64,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,32,0.004567999806669024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,64,51200,0.009720889230569204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,256,0.004987555659479565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,65536,0.042147556940714516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,51200,0.03394044438997904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,64,12288,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,128,1536,0.004271999829345279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,64,10240,0.006440889090299606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,12288,0.012094222009181976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,64,8192,0.0064399999876817065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,64,7168,0.006299555715587404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,10240,0.011067555182509951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,64,16384,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,8192,0.009584888815879822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,64,5120,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,7168,0.009336888790130615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,64,4096,0.005664888769388199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,16384,0.014464888307783337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,64,3584,0.00599644457300504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,5120,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,64,3072,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,4096,0.007328000333574083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,1024,0.009098666409651438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,64,2560,0.006333333336644703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,64,2048,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,3584,0.007279111279381647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,64,1536,0.004558222161398994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,3072,0.006679111056857639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,2560,0.006684444430801604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,64,1024,0.0035386664999855887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,2048,0.006315555423498154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,64,768,0.003555555724435382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,64,512,0.0032266666077905228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,1024,0.005626666463083691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,768,0.005617777920431561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,512,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,64,128,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,64,65536,0.010676444404655032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,64,64,0.00253866674999396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,128,0.004594666676388847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,64,32,0.002516444358560774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,64,0.004588444613748127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,32,65536,0.009006222089131674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,6144,0.008667555948098501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,64,256,0.0029279999434947968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,32,51200,0.008966222405433655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,65536,0.04179822074042427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,512,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,256,0.004963555683692296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,32,12288,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,32,10240,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,64,6144,0.006314666734801398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,12288,0.012134222520722283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,32,8192,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,1536,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,10240,0.011069333387745751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,32,6144,0.006532444308201472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,32,0.004632888982693355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,32,16384,0.006614222294754452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,32,5120,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,8192,0.009810667071077559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,51200,0.03432622220781114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,16384,0.014167999227841696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,4096,0.0074240002367231585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,32,3072,0.005972444597217772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,32,7168,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,32,2560,0.00598933340774642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,3072,0.006865777903132969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,32,2048,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,2560,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,32,1536,0.004391111019584867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,32,4096,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,32,1024,0.003903999924659729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,5120,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,32,768,0.0032986667421129015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,1536,0.005965333431959152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,32,512,0.0032524443748924467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,1024,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,32,256,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,3584,0.0069751110341813825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,32,128,0.0025333333760499954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,512,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,32,64,0.002512000087234709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,256,0.004586666822433472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,32,32,0.0025759999536805684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,128,0.004583111239804162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,64,0.004931555440028508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,32,0.004589333302444882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,768,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,51200,0.9710764355129666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,7168,0.009016888837019602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,2048,0.006268444574541516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,16384,0.3240586651696099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,16384,0.16290222273932561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,51200,0.43131110403272843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,12288,0.22775644726223418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,16384,0.13614755206637913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,51200,0.4904213481479221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,12288,0.12194666597578262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,6144,0.00867377801073922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,10240,0.1031893359290229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,16,32,3584,0.006325333482689328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,7168,0.13371200031704372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,10240,0.18949333826700845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,7168,0.06297866503397624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,12288,0.12435021665361191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,8192,0.15464888678656682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,10240,0.08688800202475654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,6144,0.11564000447591145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,6144,0.05547377798292372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,4096,0.08195377720726861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,8192,0.087044444349077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,5120,0.07094666692945692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,3584,0.06939466794331868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,4096,0.03956088754865859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,7168,0.1109422246615092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,3072,0.06056977642907036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,3584,0.03522933191723294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,5120,0.09731733136706883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,3072,0.031352887550989784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,3072,0.0444488889641232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,5120,0.047019554509056934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,2560,0.02748799986309475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,2048,0.04203288753827413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,4096,0.06039110819498698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,2048,0.023362666368484497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,2048,0.0351217786471049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,3584,0.05077066686418322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,1024,0.023583110835817125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,1536,0.026148445076412622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,6144,0.0851093331972758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,2560,0.05098844567934672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,768,0.018952889574898612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,1024,0.01538133290078905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,512,0.013779555757840475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,768,0.013328889177905189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,1024,0.01959288948112064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,256,0.008401778009202745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,768,0.01516177753607432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,512,0.012416889270146688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,128,0.006572444405820634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,256,0.010785777535703448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,256,0.009684444301658207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,1536,0.0192284451590644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,64,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,32,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,64,0.010079111489984725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,32,0.010031999813185798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,512,0.012809777425395118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,1536,0.03267644511328803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,128,0.010880000061459012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,128,0.008032889001899296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,65536,1.115581300523546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,8192,0.11132889323764378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,51200,0.8460551367865669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,16384,0.28230489624871147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,51200,0.3913102149963379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,16384,0.13104711638556585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,12288,0.2128151125378079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,65536,0.43778665860493976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,65536,0.5148497687445747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,16384,0.11123111512925889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,12288,0.08357422219382392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,51200,0.3339448769887288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,8192,0.14042222499847412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,10240,0.08326222499211629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,7168,0.12326578299204509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,10240,0.17839554945627847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,8192,0.07121244404051039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,6144,0.10744977659649319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,7168,0.08171111345291138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,7168,0.0525146656566196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,5120,0.09038044346703424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,6144,0.06822666856977674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,6144,0.05288177728652954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,5120,0.03878044419818454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,12288,0.09832000070148045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,2560,0.040489776266945735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,3584,0.06454844607247247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,4096,0.032695998748143516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,4096,0.04998133248753018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,8192,0.08864355749554104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,3584,0.029436445898479883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,3584,0.04370577798949348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,2560,0.047629333204693265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,3072,0.02642400066057841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,3072,0.036292443672815956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,2560,0.023015111684799194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,2560,0.03287911083963182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,4096,0.07230577866236369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,2048,0.01959022217326694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,2048,0.028182221783532038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,10240,0.06982400019963582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,1024,0.021316443880399067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,1536,0.016657777958446078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,1536,0.02186311119132572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,3072,0.05650133556789822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,768,0.01632000009218852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,1024,0.013441777891582914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,512,0.011728888584507836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,768,0.012072888513406118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,2048,0.038703110482957624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,768,0.013809777796268463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,512,0.011115555961926779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,512,0.011051555474599203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,1536,0.030200001266267564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,128,0.0042151109211974675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,256,0.009701333112186855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,64,0.003539555602603488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,128,0.00959022260374493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,32,0.003528888854715559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,1024,0.016680000556839835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,64,0.009022222624884712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,256,0.007520000139872233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,65536,0.325346655315823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,5120,0.05680533250172933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,65536,0.17301689253913033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,65536,0.1597742239634196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,256,0.008391111261314815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,128,0.007327111230956183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,51200,0.12422400050693089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,16384,0.042259554068247475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,16384,0.04775644342104594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,10240,0.0531582236289978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,12288,0.033424887392256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,32,0.00905777762333552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,8192,0.043812443812688194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,51200,0.25204534000820583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,8192,0.03033422099219428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,51200,0.13138400183783636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,7168,0.03802400165134006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,12288,0.07163644499248929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,6144,0.033936887979507446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,7168,0.022330666581789654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,12288,0.036474668317370944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,7168,0.031209776798884075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,6144,0.020169778002632987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,10240,0.02884355518552992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,6144,0.026597332623269822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,4096,0.02527644402450985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,5120,0.017881777551439073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,16384,0.09140621953540379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,8192,0.03583822316593594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,4096,0.015524443652894763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,3072,0.01868888901339637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,3584,0.014816888504558139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,3584,0.01906222270594703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,3072,0.013774221969975365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,2560,0.015817777978049386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,3072,0.015875554747051664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,2048,0.013205333716339536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,2560,0.012411555482281579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,10240,0.03160711129506429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,1536,0.010784889260927836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,2048,0.011063111325105032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,5120,0.02403733299838172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,3584,0.02092711130777995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,1536,0.009531555904282464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,4096,0.020815110868877836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,768,0.00692622239391009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,1024,0.008319999608728621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,1024,0.009278222090668148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,512,0.005124444348944558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,768,0.007666666474607255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,512,0.006640888750553131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,256,0.003244444520937072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,2048,0.013426666458447775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,1024,0.008350222474998897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,128,0.0028808888875775864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,256,0.006960888703664144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,256,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,64,0.0028560000161329904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,128,0.007009777757856581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,32,0.002924444360865487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,128,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,64,0.006267555471923616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,65536,0.2549537817637126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,65536,0.15197243955400255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,1536,0.011591110792424945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,65536,0.13700444168514675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,51200,0.1998071140713162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,16384,0.06718222300211589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,51200,0.11311377419365777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,5120,0.02916533417171902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,51200,0.10489600234561497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,12288,0.05252888798713684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,2560,0.014278221461508008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,16384,0.041604446040259466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,12288,0.034103999535242714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,10240,0.027758222487237718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,10240,0.025859556264347498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,768,0.007632888853549957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,512,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,8192,0.030963556634055242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,7168,0.033047109842300415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,8192,0.026874666412671406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,12288,0.0318213336997562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,6144,0.029121776421864826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,7168,0.0287715560860104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,6144,0.019865777757432725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,6144,0.02661955522166358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,32,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,5120,0.020678222179412842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,5120,0.016215999921162922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,10240,0.04515555500984192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,7168,0.020249777370029025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,3584,0.017432888348897297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,4096,0.014816888504558139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,4096,0.019303111566437613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,3072,0.015557333827018738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,3584,0.01367644468943278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,5120,0.025387555360794067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,3584,0.016568889220555622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,3072,0.012714666624863943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,16384,0.038055999411476985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,3072,0.014608888162506951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,2560,0.01144444445768992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,2560,0.013060444758998023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,4096,0.020601777566803824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,2048,0.012472000386979846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,1536,0.008692444198661381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,1536,0.010223999619483948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,8192,0.03788177834616767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,1024,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,1024,0.008237332933478886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,2560,0.013218666944238873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,512,0.004908444566859139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,768,0.007278222176763747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,2048,0.011740444435013665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,256,0.0035413333939181436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,512,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,1536,0.009392889009581672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,2048,0.009963555468453301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,128,0.0032177778581778207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,256,0.006641777853171031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,64,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,256,0.005608888963858287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,32,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,128,0.0053173332578606075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,64,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,65536,0.21818133195241293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,32,0.00589155571328269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,768,0.007037333316273159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,65536,0.14195999834272596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,512,0.006331555545330048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,65536,0.12652977307637533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,16384,0.06017333269119263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,1024,0.007897777689827813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,51200,0.10760444402694702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,12288,0.04710222283999125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,16384,0.03648622168434991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,16384,0.04014577799373203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,768,0.006361777583758037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,12288,0.03035199973318312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,8192,0.03256533212131924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,10240,0.024830222129821777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,10240,0.02678666677739885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,7168,0.02938577863905165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,8192,0.023642667465739783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,8192,0.025632888078689575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,6144,0.026151110728581745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,7168,0.01955111159218682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,128,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,7168,0.023883556326230366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,51200,0.1013431085480584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,6144,0.01770755483044518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,10240,0.03894044293297662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,6144,0.019312888383865356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,5120,0.01572800013754103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,5120,0.01796800063716041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,4096,0.013819555441538492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,4096,0.015659555792808533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,3584,0.012527999778588613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,3584,0.014466666513019137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,3072,0.012143111063374413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,3072,0.012655110822783576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,2560,0.01239911135700014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,5120,0.022701332966486614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,2048,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,4096,0.01834488908449809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,2560,0.01074399964676963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,12288,0.028825776444541082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,1536,0.009036444127559662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,2560,0.011426666544543372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,2048,0.009379555781682333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,1024,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,2048,0.01071911139620675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,51200,0.17417066627078584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,768,0.005559999909665849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,1536,0.009385777844323052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,1024,0.007325333025720384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,1024,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,768,0.007343110938866933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,768,0.006934222247865465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,256,0.0038746665749284956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,512,0.00701422244310379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,512,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,128,0.0034248888906505373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,256,0.006792888873153263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,256,0.005587555468082428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,3584,0.016166221764352586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,128,0.006790222393141852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,32,0.0029191111938820947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,1536,0.008392000363932716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,3072,0.014724445011880664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,64,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,32,0.005986666513813867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,51200,0.13211999999152288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,65536,0.07651378048790826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,51200,0.058719999260372586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,16384,0.04600355691379971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,512,0.004909333255555895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,16384,0.027267555395762127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,12288,0.03298133280542161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,16384,0.022568888134426538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,12288,0.018597332967652213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,12288,0.021401777863502502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,64,0.002874666617976295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,8192,0.025224889318148296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,10240,0.016530666086408827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,10240,0.01900533338387807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,7168,0.021653332644038733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,8192,0.01480977733929952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,65536,0.17535555362701416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,8192,0.020270221763186984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,7168,0.013727111121018728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,7168,0.018188445104493033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,5120,0.016572443975342643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,6144,0.0138586668504609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,10240,0.027782221635182697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,4096,0.014127110441525778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,5120,0.011730666789743634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,51200,0.0691208905643887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,5120,0.014007111390431723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,128,0.005366222312053044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,4096,0.010648888846238455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,3072,0.01128888875246048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,4096,0.012424888710180918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,3584,0.011591999895042844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,2560,0.010028444230556488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,3072,0.010751999914646149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,2560,0.008526222573386298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,2560,0.009767110976907942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,2048,0.00851022203763326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,2048,0.008008889026112026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,2048,0.008816888762844933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,3584,0.009678222239017487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,3072,0.00923555592695872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,1024,0.0052355556852287715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,65536,0.08937778075536092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,1024,0.006725333217117522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,768,0.004222222086456087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,1024,0.006663110935025745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,3584,0.012485333614879184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,512,0.0035742223262786865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,1536,0.0074373334646224976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,512,0.0063075555695427795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,256,0.0032222221295038858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,512,0.005921777751710679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,256,0.005991111199061076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,128,0.0031946665710873077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,6144,0.01651733285850949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,128,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,768,0.006641777853171031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,768,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,32,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,64,0.005626666463083691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,1536,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,32,0.005586666779385672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,256,0.0053182223604785065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,65536,0.08433689011467828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,65536,0.06654577785068087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,64,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,16384,0.04199999901983473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,51200,0.06291288799709745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,16384,0.021935110290845234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,6144,0.01902933253182305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,16384,0.025967111190160114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,65536,0.14986666043599448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,10240,0.028389334678649902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,12288,0.017863111363516915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,12288,0.01999288962946998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,51200,0.11775022082858616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,10240,0.015885333220163982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,10240,0.01720622181892395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,7168,0.020599999361568026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,51200,0.05429155627886454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,8192,0.019646222392718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,6144,0.018038221531444125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,7168,0.01314133322901196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,1536,0.007693332930405934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,12288,0.03325155708524916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,5120,0.015622221761279635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,6144,0.012248888611793518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,6144,0.015503111812803479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,4096,0.013440888788965015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,5120,0.011160888605647616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,5120,0.013789333403110504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,3584,0.0120666664507654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,4096,0.010055999788973067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,8192,0.014183999763594734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,3072,0.010712888505723743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,3584,0.011384888655609555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,7168,0.016855110724767048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,2560,0.009456000394291347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,2560,0.009516444471147325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,8192,0.022068444225523207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,2048,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,128,0.00498488876554701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,2560,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,4096,0.012395555774370829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,3584,0.008970666262838576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,2048,0.008370666868156856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,1536,0.00693866651919153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,1536,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,3072,0.010253333383136326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,768,0.004233777936961916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,1024,0.006602666858169768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,1024,0.006596444381607904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,512,0.003597333199448056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,768,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,256,0.0032195556494924757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,512,0.006064888917737537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,2048,0.007696888513035244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,128,0.0028897778441508612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,256,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,1024,0.005241777747869492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,64,0.0028826666788922418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,128,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,32,0.002566222308410539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,64,0.0053395554423332214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,32,0.0052728888889153796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,65536,0.12334577242533366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,1536,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,65536,0.07557333177990384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,51200,0.09753244453006321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,256,0.005250666704442766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,65536,0.06366933054394193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,16384,0.0361768901348114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,128,0.00518133325709237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,51200,0.052319112751219005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,16384,0.021031111478805542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,768,0.0064497776329517365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,16384,0.023936889237827722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,10240,0.024873778223991394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,12288,0.01809599995613098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,10240,0.015437333120240105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,10240,0.01590666671593984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,8192,0.013766222529941134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,3072,0.00869511150651508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,51200,0.05708266629113091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,7168,0.015503999259736804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,7168,0.012783110969596438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,12288,0.0173333336909612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,512,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,5120,0.015511110424995422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,8192,0.018221333622932434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,7168,0.01937244501378801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,5120,0.012734221915404001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,4096,0.011610666910807291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,12288,0.028924445311228435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,3584,0.010375111467308467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,4096,0.009134222235944536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,4096,0.011751111182901593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,6144,0.017308443784713745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,3072,0.010090666512648264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,3584,0.008418666819731394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,2560,0.008991111483838823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,3072,0.008350222474998897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,6144,0.014353776971499125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,3072,0.009697777529557547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,2560,0.007639110916190677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,2560,0.008469333251317343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,1536,0.006811555474996567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,2048,0.00793511089351442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,1536,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,1024,0.004868444469239977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,3584,0.01074222226937612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,1024,0.006302222195598815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,768,0.0039048890272776284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,2048,0.007335111498832703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,768,0.005976000179847081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,512,0.003572444534964032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,768,0.006319111006127463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,1536,0.0070293330483966405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,512,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,256,0.0032213332338465583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,512,0.005617777920431561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,128,0.0028417776856157514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,256,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,256,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,2048,0.00797333319981893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,64,0.0028684443483750024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,128,0.005906666732496685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,32,0.00282222218811512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,5120,0.01075466639465756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,64,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,65536,0.10750399695502387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,32,0.0052791109515561005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,51200,0.08570222059885661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,65536,0.0616204473707411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,8192,0.02129777769247691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,16384,0.032243556446499295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,1024,0.006599111275540457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,128,0.005126222140259213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,16384,0.020541333489947848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,16384,0.021426666114065383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,12288,0.016968000266287062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,12288,0.016502221425374348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,6144,0.011876444849703046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,10240,0.014863999353514777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,51200,0.04973333411746555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,12288,0.026245334082179602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,8192,0.016788444585270353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,7168,0.017528888252046373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,8192,0.013295110729005603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,10240,0.022247110803922016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,7168,0.012454222473833295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,7168,0.015488000379668342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,6144,0.011405333048767514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,6144,0.013657777673668332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,10240,0.014998222390810648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,4096,0.011714666253990598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,5120,0.010044444766309526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,5120,0.012429333395428128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,3584,0.010552888943089379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,4096,0.008660444782839881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,4096,0.011596444580290051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,3072,0.009672000176376766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,6144,0.01588711142539978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,65536,0.06737333536148071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,5120,0.014072888427310519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,8192,0.01980444457795885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,2560,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,2048,0.007340444458855524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,3584,0.00832177781396442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,3584,0.01052266690466139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,1536,0.005984000033802456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,2048,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,3072,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,3072,0.008979555633332994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,1536,0.006955555743641323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,1536,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,768,0.003919110943873723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,1024,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,2560,0.007673777639865875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,512,0.003563555578390757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,768,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,512,0.006131555471155379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,512,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,256,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,2560,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,128,0.0029013332807355454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,256,0.0052933332820733385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,128,0.005968889014588461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,1024,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,64,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,768,0.006264888991912206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,32,0.002557333351837264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,65536,0.08642755614386664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,64,0.0052471111218134565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,2048,0.007672000262472365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,32,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,51200,0.06832889053556654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,65536,0.06133688820732964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,1024,0.004331555631425646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,16384,0.028436445527606543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,51200,0.04773688978619046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,51200,0.05150755577617221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,12288,0.02179199953873952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,16384,0.01961777773168352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,10240,0.018966222802797954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,12288,0.01629599928855896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,128,0.004943999979231092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,12288,0.015442666080262927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,8192,0.01645955608950721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,10240,0.014507555299335055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,7168,0.014108444253603617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,8192,0.012757333616415659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,8192,0.016167110866970487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,6144,0.012434666355450949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,65536,0.05918755796220568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,256,0.005917333480384614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,5120,0.010944888823562197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,51200,0.047192000680499606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,6144,0.012778666284349231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,5120,0.009242666264375051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,16384,0.019270223047998216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,5120,0.011391999820868174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,4096,0.00835111074977451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,3584,0.008763555851247575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,7168,0.011544889046086205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,7168,0.01457155578666263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,3072,0.007963555554548899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,3584,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,3072,0.007871111234029135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,2560,0.007303111255168915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,4096,0.010382222632567087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,2560,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,2048,0.006079111248254776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,2560,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,3584,0.009365333451165093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,2048,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,1536,0.004605333424276776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,6144,0.010441777606805166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,1024,0.003891555385457145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,1536,0.00666933341158761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,4096,0.010391111175219217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,768,0.003589333345492681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,1536,0.006612444503439798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,1024,0.006476444502671559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,512,0.0032257777121331957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,1024,0.005967999911970562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,256,0.0032195556494924757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,10240,0.013883555928866068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,512,0.0052888890107472735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,128,0.0029004443850782183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,256,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,2048,0.00742222203148736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,128,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,64,0.002558222247494592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,32,0.002874666617976295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,3072,0.008011555506123437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,64,0.005274666680230035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,32,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,256,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,65536,0.05802400244606865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,51200,0.0611342191696167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,128,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,51200,0.046916445096333824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,65536,0.07832533121109009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,16384,0.018345778187115986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,768,0.005620444400442972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,16384,0.01928622192806668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,65536,0.058212445841895215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,12288,0.016176889340082806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,10240,0.01670222149954902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,10240,0.014456889695591398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,16384,0.024056888288921777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,10240,0.013448889056841532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,8192,0.01240711079703437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,768,0.00610133343272739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,8192,0.01482222146458096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,512,0.005981333139869902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,6144,0.011755555868148804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,7168,0.011348444554540845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,7168,0.014195554786258273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,5120,0.010690666735172272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,6144,0.011410666836632622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,51200,0.045144889089796275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,4096,0.00980977796845966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,7168,0.013827555709415011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,12288,0.019242667489581637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,3584,0.008336000144481659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,6144,0.009709333380063375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,3584,0.008370666868156856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,3072,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,3584,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,8192,0.014856888188256158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,2560,0.00666933341158761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,3072,0.007661333514584436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,3072,0.00775911079512702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,2048,0.00563733321097162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,2560,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,4096,0.00833777752187517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,1536,0.00462311092350218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,12288,0.01482399966981676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,1536,0.006697777658700943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,5120,0.009045333498054082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,5120,0.011160888605647616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,1024,0.0063440000845326324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,1024,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,4096,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,768,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,512,0.0031999999450312722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,1024,0.003875555677546395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,512,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,256,0.0034560000316964257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,2560,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,256,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,128,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,768,0.005582222094138463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,128,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,64,0.002562666725781229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,512,0.005570666657553778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,32,0.0026871110830042097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,64,0.005145777844720417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,768,0.003535110917356279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,65536,0.0634044475025601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,256,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,2048,0.007168889045715332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,51200,0.0511084430747562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,65536,0.057265778382619224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,128,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,1536,0.006355555521117316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,16384,0.020971554848882887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,65536,0.055777774916754834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,51200,0.04361600014898512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,12288,0.01754488878779941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,16384,0.01906844476858775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,10240,0.017204445269372728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,32,0.004910222358173794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,12288,0.014525334040323893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,8192,0.015030221806632148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,10240,0.014049778381983439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,12288,0.02027466727627648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,8192,0.011753777662913004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,7168,0.012453333371215396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,8192,0.014649778604507446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,6144,0.011048888994587792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,7168,0.01075466639465756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,51200,0.04619377851486206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,5120,0.010035555395815108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,6144,0.011406222151385413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,5120,0.009018667042255402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,4096,0.009329777624871995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,2048,0.007332444190979004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,4096,0.008362666600280339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,3584,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,7168,0.012359999948077731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,3072,0.007560888926188151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,3584,0.007914666500356462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,16384,0.017533333765135873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,2560,0.006647999915811751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,3072,0.007695111135641734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,3072,0.007240888973077138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,2048,0.005240889059172736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,2560,0.007305777735180325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,10240,0.013108444710572561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,1536,0.004567111117972268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,2048,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,4096,0.008364444805516137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,1024,0.003875555677546395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,1536,0.006671999891599019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,1536,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,3584,0.008030222521887885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,1024,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,1024,0.005625777774386936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,512,0.0032088889016045462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,5120,0.01012711144155926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,2560,0.006818666640255187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,512,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,256,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,2048,0.006317333214812809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,256,0.00591822216908137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,256,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,6144,0.009751111268997192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,64,0.002550222186578645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,128,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,32,0.0026764445420768526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,768,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,64,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,65536,0.058209776878356934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,32,0.004950222041871813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,51200,0.04538755615552267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,65536,0.05646489063898722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,65536,0.053415109713872276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,16384,0.019222221440739103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,51200,0.04546311166551378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,51200,0.041834665669335254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,12288,0.015594666202863058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,16384,0.018936889039145578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,768,0.006240889016124938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,10240,0.014496000276671516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,12288,0.014104889498816596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,512,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,10240,0.013434666726324292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,8192,0.0125608891248703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,128,0.0026897777699761917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,8192,0.011071110765139261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,7168,0.012000888586044312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,8192,0.012806221842765808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,768,0.0035537779331207275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,6144,0.01068800025516086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,128,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,7168,0.011766222616036734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,16384,0.01722400055991279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,12288,0.015625778171751235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,4096,0.00903911143541336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,5120,0.009069333473841349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,5120,0.008671999805503422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,3584,0.007864889171388414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,4096,0.008423111504978603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,4096,0.007974222302436829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,3072,0.007224000162548489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,3584,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,7168,0.01036888857682546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,3072,0.007666666474607255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,2560,0.005904888941182031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,6144,0.010028444230556488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,6144,0.011731555892361535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,2048,0.004909333255555895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,2560,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,10240,0.012759110993809171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,2048,0.006980444822046492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,2048,0.006468444648716185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,3584,0.008050666915045844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,1024,0.003936888857020272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,1536,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,1536,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,768,0.003563555578390757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,1024,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,5120,0.01036622209681405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,512,0.0033555556502607134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,768,0.006035555568006303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,2560,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,256,0.0028444443725877335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,512,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,512,0.005241777747869492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,128,0.002899555489420891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,256,0.0058959999846087555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,256,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,64,0.0025066667132907445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,128,0.004907555464241239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,32,0.0025111111915773815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,3072,0.0070159998204973005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,65536,0.05169422096676297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,32,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,65536,0.05120355553097195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,65536,0.05551111035876804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,51200,0.03723377651638455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,16384,0.01613333324591319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,128,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,51200,0.04041422075695462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,12288,0.014807111687130399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,16384,0.018407111366589863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,16384,0.01686044368478987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,64,0.004966222163703707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,10240,0.013396444420019785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,12288,0.014696000350846184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,8192,0.01240888900227017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,10240,0.012441777520709567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,1536,0.004547555413511065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,8192,0.01128533316983117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,51200,0.045010666052500405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,7168,0.010036444498433007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,6144,0.010381333529949188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,1024,0.00563822231358952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,768,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,5120,0.009349333743254343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,6144,0.008835555778609382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,10240,0.012520888613329994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,5120,0.009045333498054082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,7168,0.011056889262464313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,5120,0.008575110799736446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,3584,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,4096,0.008319999608728621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,4096,0.007669332954618666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,7168,0.010414222048388587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,3584,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,3584,0.007337777978844113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,2560,0.005625777774386936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,12288,0.013800889253616333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,3072,0.006996444529957241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,2048,0.0052675555149714155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,2560,0.007423111134105259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,2560,0.006646222124497096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,4096,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,8192,0.011409777734014722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,2048,0.006319111006127463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,1024,0.003908444609906939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,1536,0.006031999985376994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,3072,0.006618666566080517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,768,0.0035573335157500375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,1024,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,3072,0.007689778175618913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,512,0.003321777822242843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,768,0.006205333189831839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,768,0.005622222191757626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,256,0.0028479999552170434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,512,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,1536,0.0042453333735466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,2048,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,128,0.002691555561290847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,256,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,128,0.005630222045713001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,128,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,32,0.0028453332682450614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,1024,0.00564177789621883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,64,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,6144,0.009728888670603434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,32,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,51200,0.029412445094850328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,65536,0.04874933428234524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,512,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,256,0.005713777823580637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,1536,0.006611555400821898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,51200,0.03882488939497206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,12288,0.012413333687517377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,16384,0.017551110850440133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,65536,0.036855111519495644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,12288,0.01347644461525811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,10240,0.011174221833546957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,12288,0.013478221992651621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,65536,0.055063999361462064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,8192,0.01014755583471722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,10240,0.01236177815331353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,10240,0.011728888584507836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,51200,0.044534222947226636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,7168,0.009977777798970541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,8192,0.01038044442733129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,16384,0.016189333465364244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,7168,0.010211555494202508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,7168,0.009369778136412302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,6144,0.009694221946928237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,6144,0.008767999708652496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,5120,0.009008000294367472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,4096,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,64,0.0028204443968004654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,4096,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,3584,0.007280000381999546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,5120,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,3584,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,3072,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,5120,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,16384,0.014041778114106921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,2560,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,8192,0.011063999599880643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,2048,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,2560,0.007291555404663086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,6144,0.009738666315873465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,2560,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,2048,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,4096,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,2048,0.0063288890653186375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,1536,0.006665777828958299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,1536,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,768,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,3584,0.007088888850477006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,1024,0.005622222191757626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,512,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,768,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,3072,0.007648000286685095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,256,0.003168888803985384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,512,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,512,0.0052879999081293745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,1536,0.004548444516128964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,256,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,256,0.00498933345079422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,64,0.002522666628162066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,128,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,32,0.0028391112056043413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,128,0.005015111217896144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,65536,0.02643288837538825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,64,0.004965333475006951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,32,0.00490044429898262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,768,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,65536,0.054422222905688815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,65536,0.04718044400215149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,16384,0.012298666768603854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,51200,0.04378133349948459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,51200,0.03767555620935228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,12288,0.011739555332395764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,16384,0.01571733256181081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,1024,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,12288,0.012437333663304647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,51200,0.02169688873820835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,3072,0.0069751110341813825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,10240,0.011298666397730509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,128,0.0026764445420768526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,16384,0.015848888291252982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,8192,0.00981955561372969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,12288,0.01347555551264021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,10240,0.010374222364690568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,7168,0.009348444640636444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,7168,0.01017155581050449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,6144,0.00979200005531311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,8192,0.009354666703277165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,1024,0.0038515557017591265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,7168,0.009861333502663506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,4096,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,5120,0.00903200027015474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,4096,0.008343111309740279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,6144,0.008994667066468133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,3584,0.006908444480763541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,4096,0.007327111230956183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,3072,0.006317333214812809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,3584,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,10240,0.012085333466529846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,2560,0.00572533326016532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,3072,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,8192,0.011071110765139261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,2048,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,2560,0.007332444190979004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,2560,0.006608888920810487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,1536,0.00591822216908137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,2048,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,2048,0.006264888991912206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,1024,0.004582222137186262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,1536,0.005920889063013925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,6144,0.008674666285514832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,3072,0.007673777639865875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,768,0.0038853333228164246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,1024,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,5120,0.008061333662933773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,768,0.005970666805903117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,768,0.005584888988071018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,512,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,512,0.00498844434817632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,1536,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,256,0.0058222222659322954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,5120,0.008360000120268928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,256,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,1024,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,64,0.0028391112056043413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,128,0.004966222163703707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,512,0.003903999924659729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,32,0.0029013332807355454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,768,65536,0.02373955481582218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,64,0.00489777781897121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,32,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,256,0.0031991110493739447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,768,51200,0.019260444574885897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,65536,0.04666577776273092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,768,16384,0.011310222248236338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,128,0.002874666617976295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,51200,0.037512888511021934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,768,12288,0.011057777537239922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,16384,0.015217777755525378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,768,10240,0.009670221971140968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,12288,0.013439111411571503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,12288,0.012095111111799875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,768,8192,0.008579555484983657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,10240,0.012135999898115793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,65536,0.05407111181153191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,10240,0.011055111057228513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,8192,0.010913777682516309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,3584,0.007291555404663086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,51200,0.04320177767011854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,768,6144,0.008780444661776224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,7168,0.010085333552625444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,16384,0.015832000308566622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,6144,0.009687110781669617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,6144,0.00870577742656072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,768,4096,0.0074275558193524676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,5120,0.008999110923873054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,5120,0.008241777618726095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,768,3584,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,4096,0.008307555483447181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,4096,0.00742222203148736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,768,7168,0.009276444713274637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,3584,0.007963555554548899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,3584,0.007293333609898885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,768,2560,0.005618666609128316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,7168,0.009405333134863112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,128,0.005620444400442972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,768,2048,0.004939555707905027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,2560,0.007340444458855524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,2560,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,768,1536,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,2048,0.007027555671003129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,2048,0.006071110980378256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,768,1024,0.004330666528807746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,1536,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,8192,0.010006222460005019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,1024,0.006311111152172089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,3072,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,768,5120,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,768,512,0.0032293332947625052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,768,0.005980444451173146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,768,0.0052675555149714155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,768,256,0.0031840000301599503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,512,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,512,0.00526577772365676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,1536,0.006780444333950679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,256,0.004978666702906291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,768,3072,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,768,64,0.00253866674999396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,128,0.005761777775155173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,128,0.00490133340160052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,64,0.004950222041871813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,512,65536,0.017383999294704862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,32,0.004941333499219683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,65536,0.049963557057910495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,65536,0.045570668247010976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,768,128,0.0028986665937635633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,51200,0.03894311189651489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,512,16384,0.010115555591053432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,1024,0.005572444448868434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,16384,0.015855999456511605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,512,12288,0.010135999984211391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,256,0.005633777628342311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,12288,0.013447110851605734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,512,10240,0.009111999637550777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,51200,0.03701688845952352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,768,768,0.003549333247873518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,512,8192,0.00797333319981893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,10240,0.012111999922328524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,10240,0.011041777829329172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,16384,0.0147724449634552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,8192,0.010399111443095738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,512,7168,0.00906755526860555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,8192,0.009853333234786987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,512,6144,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,7168,0.010096888575288985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,512,51200,0.016197333733240765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,7168,0.009374221993817223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,6144,0.009743111001120673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,6144,0.00868977771864997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,512,4096,0.007296000089910295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,5120,0.009023110899660323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,5120,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,512,3584,0.006960888703664144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,4096,0.00832622249921163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,4096,0.00759733302725686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,512,3072,0.006641777853171031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,3584,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,3584,0.0069751110341813825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,512,2560,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,3072,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,12288,0.012096000214417776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,512,5120,0.008040889269775814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,2560,0.007334222396214803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,2560,0.006405333264006509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,512,1536,0.004581333448489507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,2048,0.007024889190991719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,2048,0.006157333238257303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,512,1024,0.003906666818592283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,1536,0.006653333289755716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,1536,0.005631999837027655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,512,768,0.003560888684458203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,1024,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,768,0.006194666855865055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,768,0.005324444423119227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,512,512,0.0032222221295038858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,512,0.006002666635645761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,768,32,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,512,0.005114666703674528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,512,128,0.002879111096262932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,256,0.005655999812814925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,256,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,1024,0.006315555423498154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,128,0.004593777573770947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,512,32,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,3072,0.006989333364698622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,64,0.004969777746333016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,256,65536,0.017119111286269296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,32,0.004606222112973531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,256,51200,0.014595554934607612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,65536,0.04680800106790331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,65536,0.043134222428003945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,256,16384,0.009451555709044138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,51200,0.03439999951256646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,128,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,256,12288,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,16384,0.014526221487257214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,512,2048,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,256,10240,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,12288,0.012135110795497894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,3072,0.007631999750932057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,10240,0.012084444363911947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,256,8192,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,10240,0.010787555740939246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,51200,0.03736977775891622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,256,7168,0.006642666541867786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,8192,0.010718222293588849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,256,6144,0.007288888924651676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,7168,0.010043555663691627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,16384,0.015847111741701763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,7168,0.009360888765917884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,256,5120,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,6144,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,256,4096,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,6144,0.008642666869693333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,5120,0.008723555339707268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,256,3584,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,4096,0.008269333177142674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,4096,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,256,3072,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,3584,0.00796622203456031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,3584,0.006928000185224745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,512,64,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,256,2560,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,3072,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,2560,0.007183111376232571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,256,2048,0.005618666609128316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,2560,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,256,1536,0.004548444516128964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,2048,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,2048,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,5120,0.008094222181373173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,256,1024,0.0035777779089079965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,1536,0.0059262220230367445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,256,768,0.003530666646030214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,1024,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,1024,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,8192,0.009744000103738572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,256,512,0.003208000005947219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,768,0.005351999981535806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,512,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,256,256,0.0028684443483750024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,512,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,256,0.005674666828579373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,256,128,0.0025351111673646504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,1536,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,128,0.005614222337802251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,256,64,0.002524444419476721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,768,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,256,32,0.002518222149875429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,64,0.00462755560874939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,12288,0.013463111387358772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,32,0.004592888885074192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,512,256,0.0028862222615215513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,128,51200,0.012520888613329994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,65536,0.04260977771547106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,128,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,51200,0.03742311067051358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,16384,0.015984000431166757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,16384,0.014483556151390076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,128,12288,0.007714666426181793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,128,65536,0.014787554740905762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,128,10240,0.0069848886794514125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,12288,0.013295999831623502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,256,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,128,8192,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,10240,0.012142221960756512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,10240,0.011025778121418424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,128,7168,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,128,16384,0.008358221915033128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,8192,0.009696000152164036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,128,6144,0.006635555376609166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,6144,0.008358221915033128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,12288,0.012087999946541257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,65536,0.045936889118618436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,6144,0.009712888962692684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,8192,0.010760000182522668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,3072,0.007636444436179266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,128,4096,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,4096,0.008076444268226624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,128,3584,0.006018666757477655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,128,5120,0.006120000034570694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,3584,0.00738933351304796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,128,3072,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,5120,0.008736888567606607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,7168,0.010247111320495605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,3072,0.007354666789372762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,128,2560,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,3072,0.007012444237867991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,128,2048,0.005981333139869902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,2560,0.007289778027269576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,4096,0.007674666742483775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,2560,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,2048,0.0069422221018208405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,2048,0.006270222365856171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,1536,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,3584,0.007968000239796108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,1536,0.005838222387764189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,128,768,0.003526222374704149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,5120,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,1024,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,7168,0.009194666312800514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,768,0.005981333139869902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,768,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,128,256,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,512,0.0057911111248864075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,512,0.004991111242108875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,128,1536,0.004557333472702238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,256,0.00564177789621883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,256,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,128,64,0.0028071111689011254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,128,1024,0.003576888806290097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,128,32,0.002749333365095986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,51200,0.03426311082310147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,64,65536,0.011564444336626263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,128,512,0.0031786666562159858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,64,51200,0.01070666644308302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,65536,0.04220799936188591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,64,16384,0.0070168889231152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,128,128,0.002867555452717675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,51200,0.033999999364217125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,64,12288,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,64,10240,0.006647999915811751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,128,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,64,8192,0.006950221955776215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,1024,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,64,7168,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,32,0.004976888911591636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,64,6144,0.006322666588756774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,7168,0.00903200027015474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,64,5120,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,6144,0.00871288859181934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,64,4096,0.005782222168313132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,128,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,64,3584,0.005618666609128316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,5120,0.007962666451931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,64,3072,0.005631999837027655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,4096,0.007342221836249034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,10240,0.010779555473062726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,64,2560,0.00598933340774642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,3584,0.006824888702895906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,64,2048,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,3072,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,64,1536,0.004561777744028303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,2560,0.006360000206364526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,64,1024,0.003549333247873518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,2048,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,64,768,0.0032177778581778207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,1536,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,64,512,0.00319377767542998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,1024,0.005247999810510212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,16384,0.014547554983033074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,64,256,0.002838222309947014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,12288,0.012115555504957834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,64,0.004944889081848992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,64,64,0.002569777684079276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,256,0.004942222187916438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,8192,0.009691555466916826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,64,0.004615111069546806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,32,65536,0.011392888923486074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,32,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,32,51200,0.01073066641887029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,65536,0.042099555333455406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,64,128,0.0025333333760499954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,32,16384,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,51200,0.034040000703599714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,32,12288,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,32,10240,0.006734222173690796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,64,32,0.002883555574549569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,16384,0.014205333259370593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,32,8192,0.006781333436568578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,12288,0.012084444363911947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,32,7168,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,10240,0.011071999867757162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,8192,0.009655111365848118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,32,5120,0.005585777676767773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,6144,0.00833155545923445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,32,4096,0.00563466673096021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,32,3584,0.006004444426960415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,5120,0.007722666694058313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,32,3072,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,4096,0.007354666789372762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,128,0.004957333207130432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,32,2560,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,32,2048,0.005632888939645555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,3072,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,32,6144,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,32,1536,0.0042160000238153665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,32,1024,0.003540444291300244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,2048,0.006258666515350342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,32,768,0.003536888708670934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,1536,0.00600355573826366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,32,512,0.0032266666077905228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,512,0.005295111073387994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,32,256,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,512,0.0052888890107472735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,3584,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,256,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,32,64,0.002570666579736604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,2560,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,128,0.004622222234805425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,32,32,0.0028320000403457214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,7168,0.009185777770148383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,64,0.004933333231343163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,768,0.005280888742870755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,32,0.004629333400064045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,768,0.0052871112194326185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,16384,0.32445422808329266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,15,32,128,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,51200,0.4893440140618219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,16384,0.16393689314524332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,51200,0.9787466261121961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,10240,0.1888586680094401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,16384,0.15982933839162192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,12288,0.12825866540273032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,8192,0.15575377146402994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,1024,0.005382222019963794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,7168,0.13371200031704372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,10240,0.0880497760242886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,8192,0.13100532690684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,8192,0.07290044758054945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,6144,0.11549599965413411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,7168,0.09461688995361328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,12288,0.22721955511305067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,7168,0.06400089131461249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,5120,0.09726399845547146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,12288,0.12173510922325982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,6144,0.08619644244511922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,5120,0.04769866665204366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,3584,0.06912977827919854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,4096,0.03944799966282315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,4096,0.06541422340604994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,3072,0.06038844585418701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,10240,0.10300711128446792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,51200,0.44864797592163086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,3584,0.05228355526924133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,2560,0.050889776812659375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,3072,0.031017776992585924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,2048,0.04187022315131294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,3072,0.04738044407632616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,2560,0.02738044493728214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,2560,0.04297777679231432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,4096,0.08179733488294813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,1536,0.032483554548687406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,2048,0.023244443866941664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,2048,0.03427377674314711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,1024,0.023614222804705303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,1536,0.01924088928434584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,768,0.019073777728610568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,1024,0.015205333630243937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,1024,0.019578667150603402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,3584,0.035030222601360746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,512,0.013506666653686099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,768,0.013451555536852943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,768,0.01586844523747762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,512,0.012772444221708508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,512,0.012523555921183692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,6144,0.06485599941677518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,5120,0.07049333386951022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,1536,0.02421333392461141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,128,0.010692444112565784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,64,0.005631999837027655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,32,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,256,0.008359111017651027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,64,0.010033778018421596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,32,0.010030222435792288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,256,0.010678222609890832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,128,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,65536,0.5144880082872179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,65536,1.099321789211697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,51200,0.8532275623745389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,128,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,65536,0.44263378779093426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,51200,0.39081690046522355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,12288,0.2088115612665812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,16384,0.1307928827073839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,10240,0.17526577578650582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,51200,0.350658655166626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,256,0.010044444766309526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,10240,0.07009600268469916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,16384,0.12740355067782932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,8192,0.08870222171147664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,8192,0.06081866555743747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,12288,0.09812711344824897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,7168,0.05145422286457486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,10240,0.08325066831376818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,16384,0.2795751094818115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,6144,0.04424000117513868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,5120,0.09080533186594646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,7168,0.12339733706580268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,5120,0.056959999932183154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,5120,0.03845244314935472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,4096,0.07267822159661187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,3584,0.06386488676071167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,4096,0.03240888979699876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,7168,0.07600177658928765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,12288,0.10257066620720758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,3072,0.05651288562350803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,3584,0.029140445921156142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,8192,0.14021510548061794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,3584,0.042488889561759106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,2560,0.047114666965272695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,3072,0.026297777891159058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,2048,0.03824977742301093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,2560,0.022984888818528917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,1536,0.029987557066811457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,2560,0.0344524449772305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,2048,0.028872887293497723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,1024,0.021307556165589228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,1536,0.02065333392884996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,4096,0.05195200112130907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,768,0.016157334049542744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,1024,0.01366311146153344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,1024,0.01668177710639106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,512,0.011555555793974133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,768,0.012137778103351593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,6144,0.06971910927030775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,768,0.013767110804716745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,256,0.007528000407748752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,512,0.01111288865407308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,512,0.010985777609878115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,2048,0.01937333411640591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,256,0.009760888914267222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,64,0.0033688888781600525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,128,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,32,0.003556444413132138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,6144,0.1055297785335117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,64,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,32,0.009023110899660323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,65536,0.31807467672559947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,3072,0.03770666652255588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,65536,0.17350133260091147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,51200,0.2498453458150228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,65536,0.14517688751220703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,51200,0.13149155510796443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,128,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,16384,0.09173066748513116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,12288,0.07151289118660821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,16384,0.04286844531695048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,16384,0.047241777181625366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,10240,0.052352888716591724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,12288,0.039035555389192365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,12288,0.036509331729676985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,8192,0.0448293321662479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,10240,0.029119110769695703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,10240,0.0317795541551378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,7168,0.03797244363360935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,8192,0.024959110551410254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,8192,0.03563377923435635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,1536,0.016580444243219163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,6144,0.03393777873780992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,7168,0.022375111778577168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,7168,0.030571556753582422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,6144,0.0199946661790212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,6144,0.0274000002278222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,4096,0.024705777565638225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,5120,0.025623111261261836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,128,0.003965333518054751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,4096,0.016226665841208566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,3584,0.02093688812520769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,256,0.008244444098737504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,3072,0.018789332773950364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,3584,0.014780445231331719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,5120,0.02884799904293484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,2560,0.015821332732836407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,3072,0.01702933344576094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,5120,0.017839999662505258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,2048,0.013445333474212222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,2560,0.015377778146002026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,4096,0.02107644412252638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,2048,0.01109955542617374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,1536,0.01090133355723487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,2048,0.012999110751681857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,1024,0.008370666868156856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,1536,0.00941244430012173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,1536,0.011038222246699862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,51200,0.1270151138305664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,1024,0.008011555506123437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,512,0.004951999833186467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,768,0.0073724447025193115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,2560,0.012060444388124677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,256,0.0032142222755485107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,512,0.007333333293596904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,512,0.006873777757088344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,128,0.0032071111102898917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,256,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,128,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,128,0.005620444400442972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,1024,0.009180444810125563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,32,0.0029253332565228143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,64,0.005963555640644497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,768,0.00776711106300354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,32,0.005963555640644497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,65536,0.25428443484836155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,256,0.00693866651919153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,51200,0.2002640035417345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,65536,0.14562133948008218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,3584,0.017788444956143696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,16384,0.06730844577153523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,51200,0.11265333493550618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,12288,0.052449779378043286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,51200,0.10916355583402847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,16384,0.04201599955558777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,10240,0.0450764430893792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,12288,0.033306668202082314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,12288,0.03180977702140808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,8192,0.037581334511439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,10240,0.025840888420740765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,10240,0.027822222974565294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,65536,0.13588178157806396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,8192,0.021853332718213398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,8192,0.02983022067281935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,6144,0.029264890485339697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,7168,0.019924443629052904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,7168,0.026661333110597398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,5120,0.025721778472264607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,6144,0.020759110649426777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,6144,0.024587555064095393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,4096,0.020617778102556866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,5120,0.016131555040677387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,5120,0.01994044416480594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,3584,0.01776266594727834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,4096,0.01421866648726993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,4096,0.019253333409627277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,3072,0.01551555593808492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,3584,0.01332622187005149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,3584,0.016043555405404832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,2560,0.013434666726324292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,3072,0.015384000208642749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,16384,0.041511999236212835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,2048,0.011714666253990598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,2560,0.011417778001891242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,2560,0.01389600005414751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,1536,0.009639999932712978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,2048,0.010029333333174387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,2048,0.012693333129088083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,1024,0.007620444728268518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,1536,0.008729777402347988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,1536,0.010122666756312052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,768,0.006261333409282897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,1024,0.007663999994595845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,1024,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,3072,0.012316444681750404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,512,0.005393777870469623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,768,0.007423111134105259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,768,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,256,0.0035422220826148987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,512,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,512,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,256,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,256,0.005608888963858287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,7168,0.03311999970012241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,128,0.00664444433318244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,64,0.0032400000426504347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,128,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,32,0.002891555635465516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,64,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,65536,0.21721155113644072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,32,0.005978666659858491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,51200,0.17315288384755453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,128,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,65536,0.12697245015038386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,16384,0.0603662199444241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,3072,0.01352266636159685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,51200,0.10758133067025079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,12288,0.04682666725582547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,16384,0.03959911068280538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,12288,0.030609776576360066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,12288,0.03144266539149814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,10240,0.03871999846564399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,65536,0.14045243793063694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,8192,0.032496889432271324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,10240,0.02492711113558875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,64,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,7168,0.029221332735485498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,8192,0.02122933334774441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,51200,0.10406577587127686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,6144,0.02611200014750163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,16384,0.04010044535001119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,5120,0.023058666123284235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,6144,0.017550221747822232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,6144,0.02036266691154904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,4096,0.018245332770877414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,5120,0.015786666009161208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,5120,0.018262222409248352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,3584,0.01607644392384423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,4096,0.013766222529941134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,4096,0.0162453336848153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,3072,0.014361777239375643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,768,0.0069351109365622205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,10240,0.026903111073705886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,8192,0.025177778469191656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,3072,0.011926222178671094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,7168,0.0196435550848643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,3072,0.013117333253224691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,2560,0.010674667027261523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,2560,0.012087999946541257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,1536,0.009003555609120263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,2048,0.009378666679064432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,2048,0.01088444474670622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,1024,0.007666666474607255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,1536,0.008396444221337637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,1536,0.008353778057628209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,768,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,3584,0.013000888956917657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,1024,0.007305777735180325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,2560,0.01236355553070704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,768,0.00674933319290479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,768,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,7168,0.022384888596004907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,256,0.0038817777401871155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,512,0.0063164445261160536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,256,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,256,0.005607999861240387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,128,0.006498666687144174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,128,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,64,0.0029004443850782183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,1024,0.007668444679843054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,32,0.002969777832428614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,3584,0.014360888136757744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,512,0.004906666775544484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,65536,0.17599556181165907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,32,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,512,0.006674666785531574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,65536,0.07044000095791288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,128,0.0032951111594835916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,51200,0.05731911129421658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,16384,0.022854222191704646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,12288,0.032815112007988825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,64,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,2048,0.010658666491508484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,51200,0.1319004429711236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,12288,0.02032355633046892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,65536,0.09319021966722275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,16384,0.04586311181386312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,10240,0.01663377715481652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,10240,0.018903111418088276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,7168,0.021893334057595994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,8192,0.014888000157144336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,8192,0.020634666085243225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,16384,0.027220444546805486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,6144,0.018980445133315194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,7168,0.017909333109855652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,6144,0.012688000169065265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,12288,0.021570665968788996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,8192,0.02548444436656104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,51200,0.06917777988645765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,4096,0.014399111270904541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,5120,0.011759111450778114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,4096,0.010808889236715106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,7168,0.013779555757840475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,4096,0.01309511148267322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,3584,0.009701333112186855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,3072,0.011494222614500256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,5120,0.01666933298110962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,3072,0.009022222624884712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,3072,0.010936888555685679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,5120,0.014524444937705994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,2048,0.00870577742656072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,2560,0.010374222364690568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,2048,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,1536,0.007556444240940942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,10240,0.027934223413467407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,6144,0.016597333881590102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,1024,0.005555555638339784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,1536,0.007296889192528195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,1536,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,2560,0.009937778115272522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,1024,0.006774222271309958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,1024,0.00666933341158761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,768,0.006450666735569636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,2560,0.008340444829728868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,768,0.006103111224042044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,256,0.0032506665835777917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,512,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,512,0.005728000154097875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,2048,0.00906577789121204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,128,0.0030248889492617715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,256,0.005311111195219888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,64,0.0025724443710512584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,128,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,128,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,768,0.004569777597983678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,64,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,65536,0.1491546630859375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,512,0.003636444608370463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,32,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,3584,0.012420444852775998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,51200,0.1174737744861179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,65536,0.08346311251322429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,51200,0.053948445452584155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,16384,0.025628444221284654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,12288,0.0336444444126553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,12288,0.019944889677895438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,16384,0.04187377625041538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,10240,0.028488887680901423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,3584,0.011412444214026133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,10240,0.015858666764365304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,8192,0.022242666946517095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,10240,0.017289777596791584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,8192,0.014158222410413953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,8192,0.01960000064637926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,7168,0.020649777518378366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,12288,0.01812177730931176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,6144,0.018223111828168232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,7168,0.01690933273898231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,6144,0.012144000165992312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,16384,0.022034666604465906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,5120,0.015496888094478183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,32,0.0028906667398081887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,4096,0.013718222578366598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,5120,0.011360888679822287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,5120,0.013818666338920593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,4096,0.010062221851613786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,4096,0.012533333566453723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,65536,0.066630224386851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,3584,0.009176889227496253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,3072,0.010666666759385003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,3584,0.010764444039927589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,2560,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,3072,0.008683555656009251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,3072,0.01071466671095954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,6144,0.015258666541841296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,2048,0.008319999608728621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,2560,0.008012444608741337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,2560,0.010046222143703038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,2048,0.007347555624114142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,2048,0.008651555412345463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,3584,0.011824888487656912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,1024,0.005249777601824866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,1536,0.006955555743641323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,768,0.0040755557517210645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,7168,0.013110222087966071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,1024,0.006597333484225803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,512,0.003581333491537306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,768,0.005974222388532426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,512,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,256,0.003220444545149803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,1536,0.007000000112586551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,256,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,1024,0.006608888920810487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,128,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,256,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,768,0.006344888773229387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,64,0.0028560000161329904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,128,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,32,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,128,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,512,0.005665777872006099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,65536,0.12318933010101318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,64,0.0053839998112784485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,32,0.005296000176005893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,51200,0.0974044468667772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,65536,0.06378577815161811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,16384,0.035664889547559954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,256,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,51200,0.05180266830656263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,51200,0.06292266978157891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,12288,0.02858933475282457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,16384,0.02107288936773936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,10240,0.02482844392458598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,12288,0.017430222696728177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,12288,0.017836444907718234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,1536,0.007075555622577667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,8192,0.021663111117151048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,10240,0.015815999772813585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,7168,0.019293333093325298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,8192,0.013800889253616333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,8192,0.018566222654448617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,6144,0.01722400055991279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,7168,0.012771555946932899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,51200,0.05697155661053128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,5120,0.015390222271283468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,6144,0.014896889527638754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,5120,0.010834666589895884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,10240,0.015533333023389181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,65536,0.07673777474297418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,4096,0.012101333174440594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,7168,0.015993777248594496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,6144,0.011735999749766456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,3072,0.010038221875826517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,3584,0.010423999693658618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,3072,0.007977777885066139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,2560,0.008999110923873054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,3072,0.010195555786291758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,2560,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,2048,0.007972444097201029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,2560,0.009304000271691216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,3584,0.010297777752081553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,1536,0.0069324444565508105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,2048,0.007225777539942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,2048,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,1024,0.004824888788991504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,1536,0.006727111008432176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,5120,0.012888888518015543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,1024,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,768,0.003887999802827835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,1024,0.006274666637182236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,512,0.003563555578390757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,768,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,768,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,256,0.0032293332947625052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,512,0.005614222337802251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,256,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,256,0.0053386667536364655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,1536,0.006656888872385025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,128,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,4096,0.011399110986126794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,64,0.0028880000528362063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,128,0.005099555684460534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,32,0.002601777720782492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,65536,0.10761244429482354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,4096,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,64,0.0052968888647026485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,512,0.005988444305128521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,51200,0.08528621991475423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,65536,0.06134400102827284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,65536,0.0672888888253106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,3584,0.008976000050703684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,51200,0.05132444368468391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,12288,0.02641333308484819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,16384,0.02033333314789666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,16384,0.021560889151361253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,10240,0.02237866653336419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,12288,0.01688088807794783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,12288,0.016524444023768108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,8192,0.019881778293185763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,10240,0.015050666199790107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,32,0.0053084443012873335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,10240,0.014902222487661572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,7168,0.017525333497259352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,8192,0.01349955548842748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,6144,0.01569600072171953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,16384,0.03236177894804213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,16384,0.02367111047108968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,51200,0.04956355690956116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,5120,0.013793778088357715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,6144,0.011376000112957425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,6144,0.014455111490355598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,5120,0.009778666827413771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,5120,0.0120666664507654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,4096,0.011760888828171624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,3584,0.010481778118345471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,8192,0.01755466726091173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,7168,0.012482666307025485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,128,0.0028684443483750024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,3072,0.009334222310119206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,3584,0.008294222255547842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,2560,0.008663111262851292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,4096,0.012072000238630505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,3072,0.00903555585278405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,2560,0.007330666813585494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,2560,0.00850044439236323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,4096,0.008894222478071848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,1536,0.006037333359320958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,2048,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,2048,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,1024,0.004280000097221798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,1536,0.0069751110341813825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,3584,0.009910222556855943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,768,0.0038702223036024305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,1024,0.00639111093348927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,1024,0.006011555592219035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,3072,0.007979555262459649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,768,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,768,0.005624889085690181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,2048,0.007527111305130853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,512,0.005971555494599872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,512,0.005544888890451855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,128,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,256,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,64,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,1536,0.00664444433318244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,128,0.00507733349998792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,32,0.0025600000388092464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,512,0.0035377778112888336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,64,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,256,0.0031999999450312722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,51200,0.06752889023886786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,65536,0.06013688776228163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,256,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,16384,0.027160889572567407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,51200,0.04770400126775106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,128,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,51200,0.04697155621316698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,12288,0.021874666213989258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,16384,0.019673777951134574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,10240,0.019098665979173448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,12288,0.016249777542220224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,12288,0.015265777707099915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,32,0.0053084443012873335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,8192,0.016185777054892648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,10240,0.013796444568369122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,7168,0.013996443814701505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,8192,0.012883555557992725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,8192,0.016333333320087857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,6144,0.012334222594896952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,7168,0.011717333561844297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,7168,0.014150222142537435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,5120,0.01075466639465756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,16384,0.01925155520439148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,6144,0.013401777380042605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,4096,0.010065777434243096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,5120,0.009249777429633671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,5120,0.011407111254003314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,3584,0.009015111459626092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,4096,0.00832088871134652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,4096,0.010766222245163388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,3072,0.008021333151393467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,65536,0.05938400162590874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,3584,0.008707555631796518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,2560,0.007000000112586551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,3072,0.008360889222886827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,6144,0.010703110860453712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,2048,0.005972444597217772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,2560,0.007359111474619971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,2560,0.008020444048775567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,1536,0.0047146665553251905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,2048,0.00701155596309238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,65536,0.08668977684444851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,10240,0.014520888527234396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,1536,0.006647999915811751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,1536,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,3584,0.008346666892369589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,1024,0.006351111249791251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,1024,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,512,0.003216000066863166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,768,0.00611111107799742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,768,0.005607999861240387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,256,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,512,0.00598933340774642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,512,0.005622222191757626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,128,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,256,0.005995555470387141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,256,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,64,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,128,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,768,0.003555555724435382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,64,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,65536,0.07816711399290296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,32,0.005067555440796746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,65536,0.057782219515906445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,2048,0.007175111108356052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,128,0.005995555470387141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,51200,0.06086666716469658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,65536,0.05810488594902886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,16384,0.02365422248840332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,3072,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,7168,0.014843554960356818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,16384,0.019554666346973844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,16384,0.018226666582955253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,12288,0.015849777393870883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,10240,0.016875555117925007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,1024,0.0038968887594011095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,32,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,10240,0.013411555853154926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,51200,0.04696444339222378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,8192,0.012452444268597497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,12288,0.019142222073343065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,7168,0.013777777552604675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,7168,0.011418666276666852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,12288,0.014793778459231058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,7168,0.013815999858909182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,6144,0.010299555957317352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,10240,0.014477334088749357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,5120,0.010399999717871347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,4096,0.00960799968904919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,51200,0.045251554912990995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,5120,0.011069333387745751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,4096,0.008378666308191087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,4096,0.01013422260681788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,3584,0.008052444292439355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,3584,0.008170666793982187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,6144,0.01309511148267322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,5120,0.009027555584907532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,3072,0.007633777956167857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,3584,0.008431110945012834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,2560,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,2048,0.005896889087226655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,3072,0.007754666937722101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,2048,0.007337777978844113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,2048,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,3072,0.008031110796663497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,1536,0.006681777950790193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,1024,0.003916444463862313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,8192,0.014813333749771118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,2560,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,768,0.0038622220357259116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,6144,0.011953777737087674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,768,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,512,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,8192,0.014754666222466363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,2560,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,256,0.0032088889016045462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,1024,0.006299555715587404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,1024,0.006098666538794835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,128,0.0029066666546795103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,256,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,128,0.005657777604129579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,768,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,64,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,128,0.004969777746333016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,32,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,65536,0.06301866637335883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,1536,0.005962666538026597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,32,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,51200,0.05109777715471056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,65536,0.05720800161361694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,65536,0.055029332637786865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,256,0.005029333134492238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,16384,0.020841777324676514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,51200,0.046190222104390465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,12288,0.01719199948840671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,1536,0.004596444467703502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,16384,0.017543110582563613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,10240,0.016794666647911072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,12288,0.01585155559910668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,512,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,512,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,10240,0.013060444758998023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,8192,0.012039110892348819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,8192,0.014882667197121514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,7168,0.012576888832781049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,7168,0.010374222364690568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,6144,0.011182222101423474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,12288,0.014476444986131458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,6144,0.011477332976129321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,5120,0.010032888915803697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,6144,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,5120,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,4096,0.009504000345865885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,16384,0.018966222802797954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,7168,0.012411555482281579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,4096,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,4096,0.008350222474998897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,3584,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,64,0.00507733349998792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,3072,0.007631111476156447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,3584,0.007672000262472365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,2560,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,3072,0.007699555820888943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,3072,0.007003555695215861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,5120,0.00942488925324546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,2048,0.005550222264395819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,2560,0.007349333001507654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,2560,0.007276444799370236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,1536,0.004587555511130227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,2048,0.007072000039948358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,2048,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,1536,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,1536,0.006051555689838197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,1024,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,1024,0.005619555711746216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,768,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,3584,0.008070222205585903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,768,0.006049777898523543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,512,0.0032115555885765287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,1024,0.0038977778620190094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,256,0.0032035555276605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,512,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,512,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,256,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,768,0.003887999802827835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,64,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,128,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,128,0.004907555464241239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,51200,0.04348177711168925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,64,0.004935111022657818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,65536,0.057849778069390186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,32,0.004958222309748332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,51200,0.045528001255459256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,65536,0.056555555926428906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,65536,0.052547554175059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,16384,0.01924977699915568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,51200,0.045280890332327954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,256,0.004938666605287128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,12288,0.015452444553375244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,16384,0.018931556079122756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,16384,0.017124444246292114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,10240,0.014484445254007975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,12288,0.015535111228624979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,12288,0.014126222994592456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,8192,0.012634666429625617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,10240,0.01351288871632682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,10240,0.012673777838548025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,8192,0.01514311134815216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,8192,0.011804444094498953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,128,0.0028524444335036804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,8192,0.01274311128589842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,6144,0.011054221954610614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,7168,0.010422222316265106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,7168,0.012078222301271228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,6144,0.009711110757456886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,32,0.002562666725781229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,6144,0.010205333431561789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,5120,0.009018667042255402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,5120,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,3584,0.007630222373538547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,4096,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,4096,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,3072,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,10240,0.01406044430202908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,3584,0.007691555553012424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,2560,0.005588444570700328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,3072,0.007804444266690149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,3072,0.007165333463086023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,5120,0.010092444717884064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,2048,0.004906666775544484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,2560,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,4096,0.008664888640244802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,2048,0.0070168889231152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,2048,0.006715555571847492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,1024,0.003912000192536248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,1536,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,3584,0.008000888758235509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,7168,0.011916444533401065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,768,0.003656889001528422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,768,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,768,0.006132444573773279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,1536,0.004563555535342959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,1536,0.00685511115524504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,512,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,256,0.0048942222363419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,1024,0.006274666637182236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,128,0.002867555452717675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,256,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,2560,0.007418666448858049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,512,0.0032213332338465583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,64,0.0028817777832349143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,32,0.0025475554996066624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,128,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,65536,0.05190044310357836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,512,0.005157333281305102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,32,0.004958222309748332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,51200,0.03712088863054911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,65536,0.055451555384529956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,65536,0.05043377810054355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,16384,0.016543999314308167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,51200,0.04485333296987745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,16384,0.01851377718978458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,128,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,12288,0.0148053334818946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,64,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,256,0.0028808888875775864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,10240,0.013280889226330651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,12288,0.013790222505728403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,10240,0.012433778080675336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,51200,0.04177333248986138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,10240,0.012448000411192576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,51200,0.04036888811323378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,7168,0.011176889141400656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,8192,0.01107911103301578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,16384,0.01668444441424476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,6144,0.010384000009960597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,7168,0.01035377797153261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,6144,0.009733333355850643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,12288,0.014505777094099255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,1024,0.005658666706747479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,8192,0.012435555458068848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,5120,0.008307555483447181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,5120,0.00904888908068339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,8192,0.011768889096048145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,4096,0.008388444781303406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,4096,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,3584,0.008017777568764156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,3072,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,6144,0.009015111459626092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,2560,0.005621333503060871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,3072,0.007666666474607255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,3072,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,2048,0.004907555464241239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,4096,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,2560,0.006660444455014334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,1536,0.004354666504595014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,2048,0.007256000406212277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,2048,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,1024,0.0038488888078265717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,1536,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,1536,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,768,0.003549333247873518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,1024,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,3584,0.007342221836249034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,1024,0.00563733321097162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,768,0.006063999815119638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,768,0.005369777894682354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,2560,0.007376000285148621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,512,0.005218666460778978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,256,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,3584,0.0074444446298811175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,128,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,512,0.0031964443624019623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,64,0.0028782222006056043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,128,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,32,0.0026391111314296722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,512,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,65536,0.036531554328070745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,64,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,32,0.004908444566859139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,7168,0.010445333189434476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,51200,0.029088000456492107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,65536,0.05507022142410278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,65536,0.048801776435640126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,51200,0.04455911119778951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,5120,0.009392889009581672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,51200,0.03894222113821242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,12288,0.012725333372751871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,16384,0.01644355555375417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,16384,0.016315556234783597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,10240,0.011394666300879585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,12288,0.013743110828929476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,256,0.0031999999450312722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,12288,0.013633777697881063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,10240,0.011838222543398539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,8192,0.01107022249036365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,256,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,16384,0.014207111464606391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,7168,0.010218666659461128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,6144,0.00982488857375251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,7168,0.00960711141427358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,5120,0.008557333714432186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,6144,0.009679999616410997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,6144,0.008840888738632202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,10240,0.012410666379663678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,4096,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,7168,0.009707555174827576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,4096,0.00828888929552502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,128,0.0028684443483750024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,4096,0.007641777396202087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,3072,0.006250666661394968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,3584,0.008007110820876228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,3584,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,2560,0.005923555543025334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,3072,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,2560,0.007353777686754863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,5120,0.007991111112965478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,2048,0.005240889059172736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,2048,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,3584,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,1536,0.006067555397748947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,1536,0.00674311113026407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,5120,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,1024,0.003942222230964237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,8192,0.01016444464524587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,1024,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,768,0.0037360000941488477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,1024,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,3072,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,768,0.006264888991912206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,512,0.0032319999817344877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,768,0.005440000030729506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,256,0.0028924445311228433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,512,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,8192,0.010427555276287926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,512,0.0053031109273433685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,256,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,256,0.005037333402368757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,64,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,128,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,32,0.002862222285734283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,64,0.005027555757098728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,65536,0.026127111580636766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,32,0.004908444566859139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,2560,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,65536,0.05463555455207825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,128,0.0029013332807355454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,16384,0.011735111474990845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,51200,0.04415377643373278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,51200,0.0376782218615214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,12288,0.011565333439244164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,16384,0.015791111522250705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,12288,0.01312622262371911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,10240,0.01035733355416192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,51200,0.021690666675567627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,10240,0.011083555718262991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,10240,0.012098666694429187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,16384,0.01620533400111728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,8192,0.009823111196359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,8192,0.011048888994587792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,12288,0.012473777764373355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,7168,0.01036977767944336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,65536,0.04706844356324938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,128,0.005623111294375525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,6144,0.009009777671760982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,6144,0.009352000223265754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,7168,0.009402666654851701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,5120,0.00815466625822915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,5120,0.008692444198661381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,7168,0.0094035557574696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,6144,0.008969777988062965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,3584,0.0069475554757648045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,4096,0.008348444269763099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,1536,0.004550222307443619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,3072,0.006266666783226862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,3584,0.00812266684240765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,5120,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,2560,0.0058551111982928375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,3072,0.00674311113026407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,2560,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,2048,0.006664888726340399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,4096,0.007531555162535773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,2048,0.006173333360089197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,1536,0.005574222240183089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,2048,0.006741333338949416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,3584,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,1024,0.004557333472702238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,1536,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,1536,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,768,0.004222222086456087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,1024,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,4096,0.007354666789372762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,1024,0.005595555735958948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,512,0.0035582222044467926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,768,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,256,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,512,0.005617777920431561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,2048,0.006275555739800136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,256,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,256,0.004937777916590373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,3072,0.007698666718271043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,128,0.005623999983072281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,128,0.004951999833186467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,32,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,2560,0.006500444478458828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,64,0.00490044429898262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,32,0.004938666605287128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,768,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,768,51200,0.01942488882276747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,512,0.005246222019195557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,65536,0.04610310991605123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,768,16384,0.01144177797767851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,51200,0.04352533486154345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,51200,0.03737066520584954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,768,12288,0.01108977778090371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,16384,0.015881778465376962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,16384,0.014589332871966891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,768,10240,0.009582222335868413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,12288,0.013459555804729462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,12288,0.012152888708644442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,768,8192,0.008664000365469191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,10240,0.012076444096035428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,10240,0.011063111325105032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,8192,0.00906044410334693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,128,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,8192,0.00976533359951443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,64,0.002891555635465516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,768,6144,0.008316444853941599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,768,65536,0.02383377816942003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,6144,0.008775999976529015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,768,5120,0.007682667010360294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,6144,0.009692444569534725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,65536,0.054135110643174916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,8192,0.010718222293588849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,5120,0.008023999631404877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,4096,0.008385777473449707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,4096,0.00737955586777793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,768,3584,0.00702400008837382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,3584,0.008014221986134848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,768,3072,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,3584,0.007287999822033777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,3072,0.007378666765160031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,768,2560,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,2560,0.006641777853171031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,2560,0.0070168889231152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,768,2048,0.004898666507667965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,2048,0.0069751110341813825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,768,1536,0.0052373334765434265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,768,7168,0.009373333719041612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,1536,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,1536,0.005972444597217772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,768,1024,0.0042239998777707415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,7168,0.01035733355416192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,768,768,0.0038817777401871155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,1024,0.006303999986913469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,1024,0.005584888988071018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,768,512,0.0035502223504914176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,768,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,768,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,768,256,0.0031928889867332247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,2048,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,768,4096,0.007348444726732042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,768,128,0.0029013332807355454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,256,0.005592888842026393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,256,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,768,64,0.0028444443725877335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,128,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,768,32,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,64,0.004910222358173794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,512,65536,0.017634666628307767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,32,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,512,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,3072,0.006890666567616993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,51200,0.03703466720051236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,128,0.004938666605287128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,512,16384,0.010306666294733683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,51200,0.038514667087131076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,16384,0.015936000479592215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,16384,0.014512000812424554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,65536,0.049921777513292104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,5120,0.009053332938088311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,512,10240,0.00903733323017756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,12288,0.013460444079505073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,12288,0.012095111111799875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,512,8192,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,10240,0.01204622205760744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,7168,0.009359111388524374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,10240,0.011145778000354767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,8192,0.01073155552148819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,8192,0.009706666900051964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,512,6144,0.008360000120268928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,7168,0.009390222529570261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,512,12288,0.010077333284748925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,512,5120,0.007690666450394525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,6144,0.009709333380063375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,512,0.0052782222628593445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,512,4096,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,5120,0.009109333157539368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,5120,0.008033778104517195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,512,3584,0.006683555742104848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,512,7168,0.009045333498054082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,3584,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,3584,0.007303111255168915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,512,3072,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,3072,0.007648000286685095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,512,2560,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,2560,0.006646222124497096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,4096,0.007639110916190677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,512,2048,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,2560,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,7168,0.010040000081062317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,2048,0.0069573331210348345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,2048,0.006151111175616582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,512,51200,0.016168889072206285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,1536,0.006657777975002925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,512,1024,0.003909333298603694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,1024,0.006108444597986009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,4096,0.008049777812427944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,512,768,0.003549333247873518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,768,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,512,512,0.003265777809752358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,768,0.00535022219022115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,512,1536,0.004570666700601578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,512,0.005787555542257097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,512,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,1536,0.005985777825117111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,256,0.005647111270162795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,512,128,0.0026515554636716843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,1024,0.005614222337802251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,128,0.005702222386995952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,512,64,0.002559111143151919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,64,0.004903111192915174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,6144,0.008681777450773451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,512,32,0.0025271111064487034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,256,65536,0.01719911065366533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,256,0.005079999979999331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,32,0.004613333278232151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,256,51200,0.0147706667582194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,65536,0.04484533270200094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,65536,0.04657155606481764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,256,16384,0.009365333451165093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,51200,0.037328001525666975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,51200,0.034288889831966825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,256,12288,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,16384,0.014510222607188754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,3072,0.006684444430801604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,256,10240,0.007359999749395583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,12288,0.013112000293201871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,12288,0.012096888489193387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,256,8192,0.006981333096822103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,10240,0.012124444047609964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,10240,0.01109688894616233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,8192,0.010728889041476779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,8192,0.009696889254781935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,256,6144,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,7168,0.010098666780524785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,16384,0.015864888827006023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,6144,0.00962755580743154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,512,256,0.0029075555503368378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,6144,0.008729777402347988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,256,4096,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,5120,0.008736888567606607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,5120,0.00835111074977451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,256,3584,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,256,7168,0.007055111229419708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,128,0.004912000149488449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,256,3072,0.005762666463851929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,3584,0.007010666860474481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,3072,0.007660444411966536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,256,2560,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,3072,0.006903111106819577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,2560,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,256,2048,0.00583022211988767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,2560,0.006343110981914733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,4096,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,4096,0.007320889168315464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,2048,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,3584,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,256,1024,0.003923555629120933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,1536,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,1536,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,256,768,0.00350577798154619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,1024,0.006252444452709622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,256,512,0.0032097777972618737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,1024,0.005647999958859549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,768,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,256,256,0.0028560000161329904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,768,0.005283555636803309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,512,0.005976000179847081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,256,128,0.002569777684079276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,512,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,256,0.005607999861240387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,256,1536,0.004564444637960858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,256,0.004909333255555895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,128,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,256,32,0.002504000026318762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,2048,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,128,65536,0.014907555447684394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,7168,0.009382222261693742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,64,0.004906666775544484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,128,51200,0.012460444536474017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,65536,0.04625599914126926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,256,5120,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,65536,0.042444444364971586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,51200,0.03716355562210083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,128,12288,0.007696888513035244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,16384,0.01584533353646596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,16384,0.014474666780895658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,128,10240,0.007000889215204451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,12288,0.013165333204799227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,256,64,0.002577777744995223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,128,8192,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,10240,0.012100444071822695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,128,0.004975999808973736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,128,7168,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,8192,0.009535111486911774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,128,16384,0.00831288927131229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,128,6144,0.006269333263238271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,7168,0.010068444742096795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,51200,0.033952001068327166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,128,5120,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,6144,0.009367999931176504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,6144,0.008719111482302347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,12288,0.012120000190205045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,128,4096,0.005643555687533484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,5120,0.007687110867765214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,10240,0.011049778097205691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,128,3584,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,32,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,3584,0.008032889001899296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,3584,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,3072,0.0067528887755341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,5120,0.008935111264387766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,128,2560,0.006262222097979651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,3072,0.007473777565691207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,8192,0.01073066641887029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,128,2048,0.005919111271699269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,2560,0.006603555546866522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,7168,0.009352888498041365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,2048,0.005971555494599872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,1536,0.005970666805903117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,1536,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,4096,0.00811644477976693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,2560,0.007269333634111617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,128,768,0.0035386664999855887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,1024,0.006116444451941385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,1024,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,2048,0.006990222467316522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,128,512,0.0032088889016045462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,768,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,128,256,0.0028453332682450614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,512,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,256,0.005630222045713001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,128,128,0.0028924445311228433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,128,1024,0.003559111307064692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,4096,0.007640889121426477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,128,64,0.0025600000388092464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,128,3072,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,768,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,128,32,0.002564444517095884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,64,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,512,0.005783111270931032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,64,51200,0.01072799993885888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,32,0.004589333302444882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,64,16384,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,65536,0.04240266813172234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,256,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,64,12288,0.007602666815121968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,51200,0.033604443073272705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,16384,0.014214222629865011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,12288,0.012102222277058495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,10240,0.010729778144094678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,128,0.005614222337802251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,128,1536,0.004262222184075249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,64,65536,0.011404444773991903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,64,6144,0.006325333482689328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,8192,0.009713778065310584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,7168,0.009140444298585257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,64,4096,0.005642666584915585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,6144,0.00869333330127928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,64,3584,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,64,10240,0.006672888994216919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,64,3072,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,4096,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,64,2560,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,64,8192,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,64,2048,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,128,0.00490133340160052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,3072,0.0068488890926043195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,64,1536,0.004554666578769684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,64,1024,0.0038924444880750445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,64,5120,0.006123555617200003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,2048,0.005973333285914526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,1536,0.005985777825117111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,64,512,0.003216888962520493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,1024,0.005614222337802251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,5120,0.008008889026112026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,64,256,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,768,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,64,128,0.0025617778301239014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,64,64,0.002554666664865282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,512,0.0053182223604785065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,64,32,0.002552888873550627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,128,0.004583111239804162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,32,65536,0.011115555961926779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,2560,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,32,0.004889777965015835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,32,51200,0.01036533299419615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,64,768,0.0035342222286595237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,32,16384,0.006984000404675801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,65536,0.04179555508825514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,32,12288,0.007300444775157505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,32,10240,0.006660444455014334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,51200,0.03420711225933499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,32,8192,0.006654222392373615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,16384,0.014496000276671516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,12288,0.012088889049159156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,10240,0.011051555474599203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,3584,0.006993778049945831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,32,6144,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,8192,0.009694221946928237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,32,5120,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,64,7168,0.006268444574541516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,7168,0.009036444127559662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,32,4096,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,32,3584,0.005607999861240387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,5120,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,32,3072,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,3584,0.0069848886794514125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,32,2560,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,3072,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,32,2048,0.005821333163314396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,256,0.004984000076850255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,32,1536,0.0042151109211974675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,64,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,32,1024,0.003547555456558863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,6144,0.008664888640244802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,32,768,0.0034977777136696708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,4096,0.007176888485749562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,1024,0.005581333405441708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,32,512,0.0032008888406885993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,32,256,0.0029013332807355454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,768,0.005285333428117964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,32,128,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,32,7168,0.006001777946949005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,512,0.005283555636803309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,256,0.004597333156400257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,2560,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,128,0.004950222041871813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,64,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,1536,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,51200,0.48739465077718097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,32,64,0.0028497777465316984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,51200,0.9788195292154948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,14,32,32,0.002537777854336633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,2048,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,16384,0.1633804506725735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,32,0.004599110947714912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,12288,0.2269066704644097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,12288,0.12189422713385688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,12288,0.10415733522839016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,10240,0.10324977503882514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,8192,0.15608178244696722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,10240,0.0867111086845398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,16384,0.32567911677890354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,8192,0.11105866564644708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,8192,0.07105955812666151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,6144,0.11537867122226292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,7168,0.1151484383477105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,7168,0.06332355737686157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,5120,0.09689244296815659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,16384,0.13647733794318304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,6144,0.05550133188565572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,51200,0.4126782152387831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,5120,0.0746871100531684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,5120,0.04678577846950955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,3584,0.06935466660393609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,7168,0.1337031126022339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,4096,0.039679110050201416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,3072,0.06029599905014038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,3584,0.03554666704601712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,3584,0.05005155669318306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,2560,0.05081244309743246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,3072,0.05028088887532552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,6144,0.08278044727113512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,2048,0.041484445333480835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,2560,0.037477334340413414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,1536,0.032425777779685125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,2048,0.023294222023752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,2048,0.03360355562633938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,1024,0.023856000767813787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,4096,0.06501688559850057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,1536,0.026378666361172993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,768,0.01886755559179518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,1024,0.015208888385030957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,1024,0.01957244508796268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,512,0.013425778183672162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,768,0.013470222552617392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,768,0.016188444362746347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,256,0.00831733312871721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,10240,0.18851644463009307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,4096,0.080549332830641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,128,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,256,0.010770666930410596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,1536,0.01926488843229082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,128,0.010810666614108615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,64,0.005912000106440649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,128,0.008160889148712158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,64,0.010079111489984725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,3072,0.03124888737996419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,32,0.010072000324726105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,512,0.01311555587583118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,512,0.012358222570684222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,256,0.009992000129487779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,32,0.005873777800136142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,65536,0.5138008859422472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,51200,0.39040443632337785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,65536,1.0999431610107422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,51200,0.8602888849046496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,16384,0.2858817842271593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,2560,0.03163289030392965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,12288,0.2093493276172214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,16384,0.1309351126352946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,16384,0.11120800177256267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,65536,0.44443999396430117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,51200,0.3176497883266873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,12288,0.08583111233181423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,12288,0.09808711210886638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,8192,0.1423057715098063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,10240,0.17460532983144125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,8192,0.09110933542251587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,8192,0.05986933575736152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,7168,0.05709244145287407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,6144,0.045109334919187755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,5120,0.09067200289832221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,10240,0.07156178024080065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,10240,0.08306133084826998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,7168,0.1256142192416721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,5120,0.06034577555126614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,5120,0.039098666773902044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,3584,0.06505955590142144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,7168,0.0767457750108507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,4096,0.052934222751193576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,3584,0.02941688895225525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,6144,0.06777066654629178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,2560,0.04800533254941305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,3072,0.04177688890033298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,2560,0.023059555225902136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,4096,0.07332533597946167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,6144,0.10630400313271417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,3072,0.05656533108817207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,2048,0.02868622210290697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,1536,0.03020711077584161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,2048,0.019475556082195707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,1024,0.021356445219781663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,1536,0.016584888100624084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,1536,0.022036444809701707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,768,0.016198222835858662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,1024,0.013457777599493662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,1024,0.016807110773192514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,512,0.011549333731333414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,768,0.012145777543385824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,768,0.013834666874673633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,4096,0.03300177719857957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,512,0.011134222149848938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,512,0.011000889043013254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,128,0.003891555385457145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,256,0.009717333647939894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,256,0.00832444429397583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,64,0.003576000117593341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,128,0.009750222166379293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,128,0.007287110719415877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,64,0.0090506664580769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,32,0.009045333498054082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,256,0.007328888608349695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,65536,0.3233777681986491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,3584,0.040130668216281466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,65536,0.17604622575971815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,51200,0.2519182099236382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,32,0.003626666549179289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,51200,0.13127556112077501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,2048,0.0381315549214681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,51200,0.11548977428012425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,16384,0.04206577936808268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,16384,0.04798310995101929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,10240,0.05309777789645725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,12288,0.03341777788268195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,12288,0.03648089038001166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,8192,0.04387022058169047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,10240,0.02900000082121955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,10240,0.03166844447453817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,7168,0.03863111138343811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,65536,0.15355200237698025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,8192,0.02435644467671712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,8192,0.03639111253950331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,6144,0.0340577761332194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,2560,0.030681778987248737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,7168,0.031209776798884075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,6144,0.028132445282406274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,6144,0.020113777783181932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,4096,0.02423200011253357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,5120,0.01755555636352963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,3072,0.02623288830121358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,4096,0.01982311076588101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,16384,0.09127289056777954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,3584,0.014733332726690503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,5120,0.028558221128251817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,5120,0.024142222272025213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,4096,0.015441776977645027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,3072,0.013627555635240344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,2560,0.01591999994383918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,2560,0.012065778175989786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,2048,0.013398222625255585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,2560,0.014183999763594734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,3072,0.018567111757066514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,1536,0.010832000109884473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,2048,0.01344622257683012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,3072,0.017433777451515198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,1024,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,3584,0.021260443660948012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,1024,0.009346666435400644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,3584,0.01754844519827101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,768,0.007227555745177799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,2048,0.011151110960377587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,768,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,512,0.004896000027656555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,1536,0.009369778136412302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,512,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,256,0.0034764446318149567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,512,0.006653333289755716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,256,0.006984000404675801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,12288,0.07110933462778728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,128,0.002921777880854077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,256,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,64,0.002817777709828483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,128,0.006983111302057902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,32,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,768,0.007701333198282454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,128,0.005587555468082428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,1536,0.011427555647161273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,64,0.006125333408514659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,65536,0.25384532080756295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,32,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,51200,0.19980266359117296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,1024,0.008343111309740279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,65536,0.15191733837127686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,65536,0.13530311319563124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,12288,0.05231555634074741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,51200,0.11279822720421685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,16384,0.041704889800813466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,10240,0.04492444462246365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,12288,0.031918221049838595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,8192,0.03754488958252801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,10240,0.02607022225856781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,16384,0.06686666938993666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,7168,0.032933334509531655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,7168,0.02233244478702545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,16384,0.03746133380466037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,6144,0.029077334536446467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,7168,0.02017600006527371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,12288,0.030019554826948378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,5120,0.025398222936524287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,6144,0.018406222263971966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,6144,0.023969777756267126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,10240,0.02794755498568217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,5120,0.021498666869269475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,8192,0.030639999442630347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,3584,0.01758311192194621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,4096,0.014518222875065274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,4096,0.018908444378111098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,3072,0.01572355628013611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,3584,0.013506666653686099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,3584,0.015633778439627755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,7168,0.030403554439544678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,2560,0.013125333521101209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,3072,0.016185777054892648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,2048,0.011454222102959951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,2560,0.012427555190192329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,4096,0.0203484445810318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,1536,0.009340444372759925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,2048,0.010855110983053843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,8192,0.022328888376553852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,1024,0.007718222008811102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,1536,0.008828444613350762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,1536,0.010384000009960597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,768,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,1024,0.007749333149856991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,1024,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,512,0.004570666700601578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,3072,0.012609778179062737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,768,0.0073315559162033936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,256,0.0035546666218174826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,512,0.007401777638329401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,5120,0.01645599967903561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,128,0.003216000066863166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,256,0.0070168889231152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,2048,0.012629333469602795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,64,0.0029022221763928733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,51200,0.10797688696119521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,128,0.006937777830494775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,32,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,64,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,65536,0.21832444932725695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,768,0.007342221836249034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,32,0.006096000058783426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,51200,0.17261155446370444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,2560,0.013367111484209696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,65536,0.14116888576083714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,16384,0.06004799736870659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,51200,0.10734933614730835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,256,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,12288,0.04691377613279554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,16384,0.03593688872125413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,10240,0.03884355558289422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,12288,0.02889333168665568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,12288,0.030654221773147583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,8192,0.03254933489693536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,10240,0.025269332859251235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,10240,0.026692445079485577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,7168,0.02953688965903388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,8192,0.021339555581410725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,8192,0.026296888788541157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,7168,0.019401777121755812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,51200,0.09984178013271755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,6144,0.026130666335423786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,16384,0.03977510995335049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,5120,0.02274133265018463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,6144,0.01756000022093455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,6144,0.020818667279349435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,5120,0.01584533353646596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,5120,0.01775733298725552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,512,0.006559111177921295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,65536,0.13025244077046713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,3584,0.01595288846227858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,3584,0.01310755560795466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,7168,0.024297777149412367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,3072,0.014360888136757744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,3584,0.014126222994592456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,2560,0.012335110869672565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,3072,0.011796444654464722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,3072,0.013238222234778933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,4096,0.018615111708641052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,2560,0.011852444046073489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,4096,0.014125333891974555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,4096,0.016234666109085083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,1536,0.008864888714419471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,1536,0.008459555606047312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,1024,0.0075786668393347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,1024,0.007648000286685095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,2048,0.010936888555685679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,768,0.005619555711746216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,1024,0.007668444679843054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,512,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,2048,0.009018667042255402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,2048,0.01111822244193819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,768,0.0069982219073507525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,256,0.0038728887836138406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,512,0.006786666810512543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,128,0.0031973332580592898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,512,0.0063591111037466265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,128,0.0052800000541739995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,256,0.005611555443869696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,128,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,128,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,2560,0.01051288925939136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,32,0.005587555468082428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,65536,0.17590933375888398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,256,0.006758222149478064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,1536,0.009360888765917884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,64,0.0032186667538351486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,51200,0.1317324505911933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,16384,0.04563911093605889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,51200,0.0691573354932997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,64,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,12288,0.032840887705485024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,768,0.007286222444640265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,12288,0.018612444400787354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,12288,0.021480000681347314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,65536,0.09092888567182754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,65536,0.07146844598982069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,8192,0.024882665938801233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,10240,0.016528889536857605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,32,0.0032302221904198327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,7168,0.02197511163022783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,51200,0.05730933613247342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,8192,0.020607110526826646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,6144,0.019109333554903667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,16384,0.02306933369901445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,7168,0.01826844447188907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,5120,0.016512889001104567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,6144,0.01276266657643848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,6144,0.016178665889634024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,4096,0.014378666877746582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,5120,0.01165511127975252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,5120,0.014483556151390076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,10240,0.018909333480728995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,4096,0.012426666915416718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,8192,0.014869333969222175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,3072,0.011343110766675739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,3584,0.010048000348938836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,3584,0.011747555600272285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,16384,0.028198222319285076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,2560,0.009861333502663506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,3072,0.009034666750166152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,2048,0.008356444537639618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,2560,0.008610666626029545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,10240,0.0277164445983039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,2560,0.010022222167915767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,2048,0.007716444631417592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,2048,0.00923288861910502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,1024,0.005583999885453119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,1536,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,1536,0.00757244477669398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,768,0.004597333156400257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,1024,0.006950221955776215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,1024,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,7168,0.013609777722093793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,768,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,768,0.006322666588756774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,256,0.0032506665835777917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,3072,0.011062222222487131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,512,0.0063271112740039825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,512,0.005623111294375525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,1536,0.007274666594134436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,128,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,256,0.0053288886944452924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,64,0.0028924445311228433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,3584,0.01276088919904497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,32,0.0029066666546795103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,64,0.005392888767851724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,512,0.0037671112351947357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,32,0.005647999958859549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,65536,0.14883732795715332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,51200,0.11681866645812988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,65536,0.08265955580605401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,65536,0.0661768913269043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,16384,0.04191822144720289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,128,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,4096,0.01072088877360026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,51200,0.06301244762208727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,128,0.004994666824738185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,16384,0.022044445077578228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,10240,0.0284897784392039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,12288,0.01787911189926995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,12288,0.01996977792845832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,8192,0.02184533410602146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,10240,0.01584533353646596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,10240,0.017264000243610807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,7168,0.02055288851261139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,256,0.00601333338353369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,8192,0.01998844411638048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,6144,0.017782222893502977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,7168,0.01314755529165268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,7168,0.017439999514155917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,12288,0.033462223079469465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,5120,0.015493333339691162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,6144,0.012078222301271228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,6144,0.015524443652894763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,4096,0.013783999615245394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,5120,0.011071110765139261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,4096,0.009735110733244155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,4096,0.012087999946541257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,3072,0.010799999866220685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,3584,0.01072622256146537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,3072,0.008000888758235509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,2560,0.009702222214804756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,51200,0.05418755610783895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,16384,0.02581866747803158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,3584,0.011973333027627734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,5120,0.013467555244763693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,3584,0.00904622260067198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,2048,0.008781332936551835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,1536,0.007332444190979004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,1536,0.007000000112586551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,1024,0.0053102220926019884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,2560,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,2048,0.008430221842394935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,768,0.004260444392760594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,1024,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,2048,0.007370666497283512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,768,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,8192,0.014396443963050842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,768,0.00601333338353369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,256,0.0032026666320032547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,512,0.006292444550328785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,3072,0.010846222440401712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,512,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,1536,0.007324444750944774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,256,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,64,0.002575111058023241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,256,0.0053119998839166425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,128,0.0059262220230367445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,32,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,128,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,65536,0.12253244717915852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,64,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,32,0.0053680001033677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,51200,0.09732266929414536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,512,0.0035502223504914176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,65536,0.07394311163160536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,65536,0.06372621986601087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,51200,0.05187199844254387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,12288,0.028451555305057104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,16384,0.02109688851568434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,16384,0.023426666855812073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,10240,0.024767110745112102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,2560,0.009359999663299983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,1024,0.006406222366624408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,8192,0.021581333544519212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,10240,0.015443555182880826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,10240,0.0158897770775689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,7168,0.01920622256067064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,8192,0.01367555558681488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,8192,0.018608889646000333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,6144,0.017143110434214275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,51200,0.05691111087799072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,128,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,5120,0.015261333849694995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,6144,0.01183999992079205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,6144,0.014481777946154276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,12288,0.017306667235162523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,5120,0.010445333189434476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,12288,0.018094221750895183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,5120,0.0127688886390792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,3584,0.01035733355416192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,4096,0.009364444348547194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,4096,0.012121777567598553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,16384,0.03567466802067227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,7168,0.012784889174832238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,3072,0.01015822258260515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,3584,0.01036444471942054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,7168,0.016223111086421542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,3072,0.007917332980367873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,3072,0.010421333213647207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,2560,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,4096,0.01181777815024058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,2560,0.009022222624884712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,1536,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,2048,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,1024,0.004865777575307422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,1536,0.00703288863102595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,1536,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,768,0.004210666649871402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,1024,0.006598222172922558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,1024,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,512,0.003532444437344869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,768,0.006322666588756774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,768,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,256,0.0031991110493739447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,512,0.005590222362014983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,128,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,256,0.005987555616431766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,256,0.0052471111218134565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,64,0.0028764444092909494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,128,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,32,0.0028684443483750024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,3584,0.008604444563388824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,2560,0.008984888593355814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,65536,0.1071386668417189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,2048,0.007635555333561367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,512,0.006127111199829314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,51200,0.085735109117296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,65536,0.061434666315714516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,2048,0.007970666719807519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,51200,0.04957866668701172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,128,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,12288,0.02604355580276913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,16384,0.02131822208563487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,12288,0.016887111796273124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,64,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,12288,0.016544888416926067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,32,0.005245333330498801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,8192,0.019513777560657926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,10240,0.015142222245534262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,10240,0.01481599940194024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,16384,0.03237066666285197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,7168,0.01740444368786282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,51200,0.05141511228349474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,8192,0.013409777647919126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,6144,0.015672889020707872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,7168,0.012641777594884237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,16384,0.020405333903100755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,5120,0.013873777455753751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,7168,0.01516533394654592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,4096,0.012172444827026792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,6144,0.013549333645237817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,5120,0.010104888843165504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,10240,0.022309333086013794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,4096,0.008740444150235917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,4096,0.011382222175598145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,3072,0.009564444422721863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,65536,0.067413330078125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,3584,0.009707555174827576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,2560,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,3072,0.007739555504586961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,8192,0.01884355478816562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,2048,0.0073626670572492815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,2560,0.007436444362004598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,6144,0.01145066652033064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,1536,0.006314666734801398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,2048,0.00720088928937912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,2048,0.007713778151406183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,1024,0.004273777620659934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,1536,0.006637333167923822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,1536,0.007002666592597961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,768,0.00424799985355801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,3584,0.00814577771557702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,1024,0.00630044440428416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,512,0.003539555602603488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,768,0.006322666588756774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,3072,0.009344889058007134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,256,0.003228444399105178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,512,0.006070222291681502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,2560,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,512,0.005476444545719359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,128,0.0028444443725877335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,256,0.006005333529578314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,64,0.002867555452717675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,128,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,32,0.002516444358560774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,128,0.004959110998445087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,1024,0.006351999938488007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,64,0.005296000176005893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,32,0.005426666802830166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,768,0.005987555616431766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,65536,0.059225777784983315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,65536,0.06101510922114054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,256,0.005240889059172736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,51200,0.046726223495271474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,16384,0.019682667321628995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,65536,0.08708088927798802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,12288,0.021664000219768945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,16384,0.01958755486541324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,51200,0.06780266761779785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,10240,0.019139554765489366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,12288,0.015215999550289579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,3584,0.010692444112565784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,16384,0.027684445182482403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,10240,0.013949332965744866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,8192,0.012969777815871768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,7168,0.013804444836245643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,7168,0.014521777629852295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,12288,0.016176889340082806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,6144,0.01237777786122428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,7168,0.011750222080283694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,5120,0.012464889221721225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,6144,0.010800888968838586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,8192,0.01686311099264357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,6144,0.012475555141766867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,4096,0.009688888986905416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,5120,0.011418666276666852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,8192,0.017268444101015728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,4096,0.010428444378905825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,3584,0.008174222376611497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,3072,0.008010666403505538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,51200,0.04788800080617269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,10240,0.014482667048772177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,3072,0.00831733312871721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,5120,0.009735999835862054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,2560,0.00740799970097012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,2560,0.007335111498832703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,3584,0.008441777692900764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,2048,0.006955555743641323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,1536,0.004632888982693355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,2048,0.007127111156781514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,5120,0.011078221930397881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,1024,0.0041804446114434134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,1536,0.006677333265542984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,3072,0.007688889073001013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,768,0.003580444388919406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,1024,0.006265777680608962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,1024,0.006135111053784688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,2048,0.006005333529578314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,768,0.006266666783226862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,768,0.005648889061477449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,512,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,512,0.005629333357016246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,256,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,128,0.0029022221763928733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,2560,0.007286222444640265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,1536,0.006606222026877933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,64,0.0028862222615215513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,128,0.005965333431959152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,32,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,128,0.004960888789759742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,512,0.0034222222036785553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,65536,0.0782711108525594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,64,0.005208000126812193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,256,0.0029128889242808023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,32,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,65536,0.05845955345365736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,65536,0.05748711029688517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,16384,0.02386755579047733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,51200,0.04679022232691447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,256,0.0051857779423395796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,12288,0.019147555033365887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,16384,0.019359999232821994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,10240,0.01688088807794783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,12288,0.015942222542232938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,12288,0.01498311095767551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,3584,0.008719999757077958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,8192,0.01464355488618215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,10240,0.013434666726324292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,8192,0.012536889149083031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,7168,0.013792888985739814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,51200,0.060918225182427295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,6144,0.011960888902346293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,7168,0.013942221800486246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,51200,0.045181334018707275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,5120,0.010647999743620554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,16384,0.018562666243977018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,6144,0.012413333687517377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,4096,0.009356444080670675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,5120,0.009012444151772393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,10240,0.01443555619981554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,5120,0.011690666278203329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,4096,0.008349333372380998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,4096,0.009687999884287516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,3072,0.007920888562997183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,3584,0.008001777860853408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,3584,0.008314666648705801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,7168,0.01142133358452055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,2560,0.006989333364698622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,3072,0.007984889050324758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,6144,0.010404444403118556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,2048,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,2560,0.0073831114504072405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,2560,0.007672000262472365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,1536,0.00461777796347936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,2048,0.007111111448870764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,3584,0.008468444148699442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,2048,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,1024,0.0038933331767717996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,1536,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,8192,0.014815110299322339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,768,0.0038408889538711975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,1536,0.00665244460105896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,3072,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,512,0.003237333355678452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,768,0.005624889085690181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,768,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,256,0.0032088889016045462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,512,0.005980444451173146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,128,0.0028924445311228433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,512,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,4096,0.00849955528974533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,256,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,64,0.002518222149875429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,128,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,128,0.004971555537647671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,65536,0.06392888890372382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,32,0.00508088908261723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,1024,0.006416000011894438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,1024,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,51200,0.05067911081843906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,16384,0.020620443754725985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,65536,0.05582933293448555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,51200,0.04615288972854614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,51200,0.0433919992711809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,16384,0.019268444842762418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,16384,0.01753777762254079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,12288,0.015861334072219003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,12288,0.014416888356208801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,64,0.005026666654480828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,10240,0.014010666145218743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,10240,0.013086222112178802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,8192,0.011791999969217511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,7168,0.012409778104888068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,8192,0.014147554834683737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,12288,0.01718933383623759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,6144,0.011112000379297467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,7168,0.01038755559258991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,7168,0.012110222544935016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,10240,0.016813332835833233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,5120,0.010054222411579555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,6144,0.011749332977665795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,4096,0.009142222503821055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,5120,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,8192,0.0151582227812873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,3584,0.008199111455016667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,4096,0.008379555410808986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,65536,0.05751199854744805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,3072,0.007391110890441471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,3584,0.007640889121426477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,256,0.00573955559068256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,2560,0.006642666541867786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,3072,0.007640889121426477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,32,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,6144,0.009717333647939894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,2560,0.007324444750944774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,2560,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,1536,0.004592888885074192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,2048,0.0070773329999711775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,5120,0.010030222435792288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,2048,0.006693333387374878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,1536,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,1536,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,3584,0.008262222011884054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,1024,0.005919999960396025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,768,0.0064693333374129395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,512,0.0032186667538351486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,2048,0.005199111170238919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,512,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,512,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,1024,0.0038853333228164246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,256,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,1024,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,128,0.004895110925038655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,128,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,64,0.0026319999661710528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,256,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,32,0.0025235555238193935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,64,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,768,0.003683555457327101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,65536,0.05794399976730347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,32,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,51200,0.04519199993875292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,65536,0.05630133549372355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,3072,0.007696000238259633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,65536,0.052519109514024526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,16384,0.018971555762820773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,51200,0.04552266663975186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,16384,0.017157332764731515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,12288,0.015503999259736804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,4096,0.007991111112965478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,12288,0.014140443669425117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,12288,0.0155395550860299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,10240,0.014157333307796054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,768,0.005623999983072281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,51200,0.041871110598246254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,10240,0.012749333348539142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,16384,0.018983110785484314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,8192,0.011064888702498542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,7168,0.011761777930789523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,8192,0.013112000293201871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,6144,0.011069333387745751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,128,0.0028728888266616394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,7168,0.011566222541862063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,5120,0.01036800030204985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,6144,0.010272888673676385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,4096,0.008647110727098253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,5120,0.00906311141120063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,5120,0.008661333057615492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,3584,0.007979555262459649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,4096,0.008415111237102086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,4096,0.007663110891977946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,3072,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,3584,0.008150222400824228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,3584,0.007290666302045186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,256,0.005227555417352253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,3072,0.007732444339328342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,3072,0.007096889118353526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,6144,0.009703111317422655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,2560,0.0073626670572492815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,2560,0.0069431112044387394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,1536,0.004575110971927643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,2048,0.007019555403126611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,8192,0.012812444733248817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,1024,0.003935111065705617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,1536,0.006672888994216919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,1536,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,768,0.003576888806290097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,1024,0.0063786668082078295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,1024,0.005656888915432825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,7168,0.010674667027261523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,768,0.005991999887757831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,768,0.005423111220200856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,256,0.0029155556112527847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,512,0.005976000179847081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,512,0.005239999956554837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,128,0.002699555622206794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,256,0.005924444645643234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,2048,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,256,0.004984000076850255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,128,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,128,0.004945777770545748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,64,0.005215999980767568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,2560,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,65536,0.05001599921120537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,2048,0.005247999810510212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,51200,0.03703022334310744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,65536,0.05087555448214213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,10240,0.013131555583741931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,51200,0.04483377933502197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,16384,0.016532444291644625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,32,0.002887111157178879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,16384,0.01852888862291972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,16384,0.016809778081046212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,512,0.0034088889757792154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,32,0.004940444396601783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,12288,0.014436443646748861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,12288,0.013791110780504016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,8192,0.012415111064910889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,10240,0.012436444560686747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,10240,0.012413333687517377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,7168,0.011296889020336999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,8192,0.011060444845093621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,64,0.0028693332440323303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,51200,0.04031022058592902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,6144,0.010439111126793755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,7168,0.010415111151006486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,7168,0.010196444061067369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,6144,0.009691555466916826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,6144,0.00903200027015474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,4096,0.008255110846625434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,5120,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,65536,0.055252446068657764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,4096,0.008353778057628209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,4096,0.007623111208279927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,3584,0.008057778080304464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,3072,0.006353777729802662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,5120,0.00979377743270662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,3072,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,5120,0.009012444151772393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,2560,0.005766222046481238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,2560,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,3584,0.007629333270920648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,2560,0.006584888945023219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,8192,0.012088889049159156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,3584,0.007272889216740926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,12288,0.014693333042992486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,1536,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,10240,0.013269333375824822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,1024,0.003936000168323517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,1536,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,2048,0.004975111120276981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,768,0.003539555602603488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,1024,0.00562755556570159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,2048,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,768,0.006203555398517185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,768,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,512,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,3072,0.007680888805124495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,256,0.0032035555276605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,256,0.005620444400442972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,128,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,256,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,128,0.005656888915432825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,2048,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,64,0.00250133333934678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,128,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,32,0.0025191110455327565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,65536,0.03660088777542114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,512,0.005312888986534542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,32,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,65536,0.05520266625616285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,65536,0.04859822326236301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,16384,0.013775111072593264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,1024,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,51200,0.04458222124311659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,51200,0.03870577613512675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,16384,0.017549332645204332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,16384,0.01626844373014238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,1536,0.004580444345871608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,12288,0.013529777526855469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,12288,0.013423110875818463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,8192,0.010363555616802638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,10240,0.01219466659757826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,64,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,7168,0.009691555466916826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,8192,0.011071110765139261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,8192,0.010168888502650792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,6144,0.009688888986905416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,7168,0.01037688884470198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,7168,0.009376889301670922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,12288,0.012524444195959302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,5120,0.008660444782839881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,6144,0.008782222039169734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,512,0.003265777809752358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,4096,0.0074782222509384155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,5120,0.008153777983453538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,4096,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,10240,0.011575111084514193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,51200,0.029264890485339697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,3072,0.006322666588756774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,3584,0.008010666403505538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,3584,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,2560,0.005700444595681296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,3072,0.007666666474607255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,3072,0.00701422244310379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,2560,0.007330666813585494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,2560,0.006341333190600078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,5120,0.008983111215962304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,1536,0.004567111117972268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,2048,0.007195555501514011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,3584,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,1024,0.003927111211750242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,1536,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,1536,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,768,0.003539555602603488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,1024,0.006266666783226862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,1024,0.005683555371231503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,6144,0.009679111341635386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,512,0.0032471112079090546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,768,0.0061919999619325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,10240,0.011426666544543372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,512,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,512,0.005224888937340842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,2048,0.006309333360857434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,128,0.0028391112056043413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,256,0.004942222187916438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,64,0.0026782221264309357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,128,0.005643555687533484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,32,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,64,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,65536,0.026203556193245783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,2048,0.005009777843952179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,256,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,51200,0.021644444929228887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,65536,0.05473866727617052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,65536,0.046774221791161426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,16384,0.012631110846996307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,256,0.005965333431959152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,4096,0.0075031113293435835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,12288,0.01148000028398302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,51200,0.037668443388409086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,16384,0.016148444679048326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,128,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,16384,0.015642666154437594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,12288,0.013462222284740873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,8192,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,10240,0.012086222569147745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,10240,0.011067555182509951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,8192,0.01071999967098236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,8192,0.010044444766309526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,7168,0.009628444082207149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,51200,0.04412000046836006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,7168,0.010041777458455827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,6144,0.009047110875447592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,7168,0.009360888765917884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,5120,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,10240,0.010416000253624385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,6144,0.009748444788985783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,4096,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,12288,0.012432888978057437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,5120,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,3584,0.006959111326270633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,32,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,4096,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,3072,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,3584,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,3584,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,3072,0.007634667058785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,3072,0.0069520001610120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,2560,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,2048,0.006263999889294307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,6144,0.00869777798652649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,5120,0.008019555773999956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,2048,0.0069448889957533936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,2048,0.006308444258239534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,4096,0.007367110914654202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,1024,0.004938666605287128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,1536,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,1536,0.0057813334796163775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,768,0.0042364444169733255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,1024,0.005636444522274866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,512,0.003576888806290097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,2560,0.006646222124497096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,768,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,256,0.0032257777121331957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,512,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,512,0.005285333428117964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,768,0.005249777601824866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,128,0.0029111111329661477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,256,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,64,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,128,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,32,0.0028888889484935333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,128,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,768,65536,0.023767110374238756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,64,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,768,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,65536,0.054309331708484225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,65536,0.04626489016744825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,1536,0.005646222167544895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,51200,0.03734044565094842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,256,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,2560,0.006011555592219035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,1024,0.00630044440428416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,32,0.004942222187916438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,768,51200,0.019267555740144517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,12288,0.012101333174440594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,12288,0.01346666696998808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,51200,0.04350399971008301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,10240,0.012131555212868584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,768,16384,0.011436444189813403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,10240,0.011071999867757162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,768,7168,0.00943022221326828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,8192,0.010772444307804108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,16384,0.0159288893143336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,8192,0.010096000300513374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,768,6144,0.008345777789751688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,7168,0.010395555860466426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,7168,0.009349333743254343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,6144,0.009358222285906473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,6144,0.00870488915178511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,768,10240,0.009706666900051964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,5120,0.00868444475862715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,5120,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,768,3584,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,4096,0.008011555506123437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,768,8192,0.00867377801073922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,4096,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,768,3072,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,3584,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,768,2560,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,3072,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,768,12288,0.011040888726711273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,3072,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,768,2048,0.004952888935804367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,2560,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,2560,0.006351999938488007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,2048,0.007060444189442529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,2048,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,768,1024,0.003929777691761653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,1536,0.006671999891599019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,1536,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,768,768,0.0037271111375755737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,1024,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,1024,0.005588444570700328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,3584,0.007131555842028723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,768,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,16384,0.0151573336786694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,768,5120,0.007975999679830339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,768,256,0.0029066666546795103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,512,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,768,4096,0.007348444726732042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,768,128,0.002847111059559716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,256,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,128,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,768,64,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,768,512,0.0032506665835777917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,768,32,0.0025235555238193935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,64,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,768,1536,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,32,0.004933333231343163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,256,0.005629333357016246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,65536,0.049867557154761426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,65536,0.045569777488708496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,51200,0.03854666815863715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,128,0.004959110998445087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,512,16384,0.010316444767846001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,16384,0.015903110305468243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,512,12288,0.010058666268984476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,512,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,12288,0.013449778159459433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,512,10240,0.008998221821255153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,512,51200,0.015846222639083862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,10240,0.012106666962305704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,10240,0.011069333387745751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,51200,0.03666222095489502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,512,7168,0.009322666459613377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,8192,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,16384,0.014521777629852295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,512,65536,0.0176862229903539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,7168,0.00908088849650489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,6144,0.009362666971153682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,6144,0.00871288859181934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,512,5120,0.007985778152942657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,512,8192,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,512,4096,0.007328888608349695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,5120,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,8192,0.010044444766309526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,4096,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,7168,0.010258666343159145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,3584,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,512,3072,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,5120,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,3072,0.0069848886794514125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,768,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,3072,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,512,3584,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,2560,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,512,6144,0.008356444537639618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,2560,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,512,1536,0.004599110947714912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,2048,0.0069422221018208405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,12288,0.012432888978057437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,512,1024,0.0039057777159743835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,1536,0.006643555644485686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,1536,0.00600000015563435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,512,768,0.003592888928121991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,1024,0.006334222025341458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,512,2560,0.006267555471923616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,512,512,0.003192000091075897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,768,0.006142222219043308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,4096,0.0069573331210348345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,768,0.005425777700212266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,512,2048,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,512,0.005984000033802456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,512,128,0.0028737777223189673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,512,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,256,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,3584,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,256,0.004959110998445087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,512,32,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,128,0.004945777770545748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,256,65536,0.017216889394654166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,64,0.004910222358173794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,1024,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,65536,0.04655999938646952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,256,51200,0.014817777607176038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,65536,0.042857779396904834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,256,16384,0.009395555489593083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,51200,0.03734133309788174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,51200,0.0346684455871582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,256,12288,0.008034666379292807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,2048,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,12288,0.013335111240545908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,12288,0.012256888879670037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,32,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,10240,0.011063111325105032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,10240,0.012076444096035428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,256,8192,0.006960888703664144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,512,64,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,8192,0.010091555615266165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,256,7168,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,7168,0.010039110978444418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,256,10240,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,7168,0.009232000344329411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,6144,0.009374221993817223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,6144,0.008347555167145198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,512,256,0.0031902222997612427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,8192,0.010846222440401712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,16384,0.014524444937705994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,256,4096,0.0063528890411059065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,5120,0.008980444735950893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,256,6144,0.007286222444640265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,4096,0.00832088871134652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,256,3584,0.006268444574541516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,4096,0.007357333269384172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,256,3072,0.005984888722499211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,3584,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,3584,0.007010666860474481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,256,2560,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,3072,0.006675555474228329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,16384,0.015885333220163982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,256,2048,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,2560,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,2048,0.0069448889957533936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,256,1536,0.004551110996140374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,2048,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,256,1024,0.003927111211750242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,256,5120,0.006955555743641323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,3072,0.007663999994595845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,256,768,0.003552888830502828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,1024,0.00599644457300504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,128,0.0058142224119769196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,2560,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,256,512,0.0031742221779293488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,768,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,256,256,0.0028782222006056043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,512,0.005247999810510212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,1536,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,256,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,256,128,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,128,0.005667555663320754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,256,64,0.00254488881263468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,1024,0.005653333332803514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,5120,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,256,32,0.002518222149875429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,128,65536,0.015169777803950839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,768,0.005476444545719359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,32,0.004609777695602841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,128,51200,0.012444444828563266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,65536,0.04619822237226698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,65536,0.042408890194363065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,256,0.005646222167544895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,1536,0.005994666781690385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,51200,0.033887111478381686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,16384,0.015861334072219003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,128,12288,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,128,0.004585777719815572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,128,10240,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,12288,0.013305777476893531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,12288,0.01223111069864697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,64,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,10240,0.010891555911964841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,8192,0.010442666709423065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,128,16384,0.00832622249921163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,128,7168,0.006304889089531369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,51200,0.037343111303117536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,128,6144,0.006240889016124938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,7168,0.010063110954231685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,16384,0.014506666196717156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,7168,0.009413333402739631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,6144,0.009348444640636444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,6144,0.008685333033402761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,128,4096,0.0058311112225055695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,5120,0.00868000007337994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,512,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,5120,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,4096,0.008115555677149031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,4096,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,128,3072,0.00592533333433999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,8192,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,3584,0.007341333561473423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,128,2560,0.0063573333124319715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,3072,0.007632888853549957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,3072,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,128,5120,0.005975111077229182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,2560,0.007334222396214803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,2560,0.006385777973466449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,2048,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,128,8192,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,2048,0.006268444574541516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,128,3584,0.006008000009589725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,128,1024,0.003593777616818746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,1536,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,128,768,0.0033937777496046494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,1536,0.005702222386995952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,1024,0.005583111196756363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,768,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,768,0.005624889085690181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,128,2048,0.005889777921968036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,128,256,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,512,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,128,1536,0.004605333424276776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,10240,0.011755555868148804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,128,128,0.0025555555605226094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,256,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,128,0.00573333352804184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,1024,0.006264888991912206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,128,0.004583999928500918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,128,32,0.002528888897763358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,128,512,0.00318755561278926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,64,65536,0.011402666568756104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,64,0.00490399988161193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,64,51200,0.010763555765151978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,32,0.004585777719815572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,512,0.006011555592219035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,64,16384,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,64,12288,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,65536,0.04214844438764784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,256,0.005595555735958948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,51200,0.03398222062322829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,64,8192,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,12288,0.011909333368142446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,128,64,0.00254844439526399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,64,7168,0.005968889014588461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,64,6144,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,7168,0.009454222189055549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,64,5120,0.005894222193294101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,6144,0.008563555777072906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,64,4096,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,64,10240,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,64,3584,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,16384,0.014492443866199918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,64,3072,0.005965333431959152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,3584,0.007903111477692923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,64,2560,0.005921777751710679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,3584,0.0070159998204973005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,8192,0.009755555954244401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,64,2048,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,3072,0.006710222197903528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,2560,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,2048,0.006133333262470033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,1536,0.005973333285914526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,5120,0.008062221937709385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,1024,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,64,512,0.003585777762863371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,4096,0.0069875551594628235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,64,256,0.0028782222006056043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,64,128,0.0028248888750871024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,10240,0.011057777537239922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,64,64,0.0025155554629034465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,256,0.0047102222839991255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,64,32,0.0025395556456512874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,128,0.004958222309748332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,32,65536,0.011476444701353708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,64,1024,0.0035546666218174826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,32,51200,0.010420444111029306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,32,0.004589333302444882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,32,16384,0.006647999915811751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,65536,0.04224888814820183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,32,12288,0.006691555596060223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,16384,0.014163555370436775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,32,10240,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,64,1536,0.004271111140648524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,12288,0.012246222131782107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,32,8192,0.006612444503439798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,32,7168,0.00655377780397733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,64,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,32,6144,0.006275555739800136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,8192,0.009720889230569204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,32,5120,0.005635555419656966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,7168,0.009020444419648912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,32,4096,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,6144,0.008374222450786168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,32,3584,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,5120,0.008217777642938826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,512,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,4096,0.007343110938866933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,32,2560,0.006010666489601135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,3584,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,32,2048,0.005608888963858287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,10240,0.010737777584128909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,32,1536,0.004589333302444882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,64,768,0.00351555562681622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,32,1024,0.0036657779581016968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,2048,0.006235555642180973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,32,768,0.0035377778112888336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,1536,0.005647999958859549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,51200,0.03360889024204678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,32,512,0.0032195556494924757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,1024,0.00555377784702513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,32,256,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,32,128,0.0025395556456512874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,32,3072,0.005616888817813661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,32,64,0.002558222247494592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,512,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,13,32,32,0.00253866674999396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,128,0.004590222405062782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,2560,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,32,0.0047591109242704184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,768,0.00564177789621883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,768,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,51200,0.48832088046603733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,256,0.004957333207130432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,3072,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,16384,0.324111991458469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,64,0.004946666873163647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,16384,0.16243110762702093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,12288,0.22671466403537324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,16384,0.13722933663262263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,12288,0.12159910466935898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,51200,0.453729788462321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,10240,0.1885626713434855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,10240,0.12603910764058432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,16384,0.17048888736300996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,12288,0.12787377834320068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,51200,0.9670755598280164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,12288,0.10328000121646458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,10240,0.10911199781629775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,8192,0.11279377672407363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,7168,0.13343200418684217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,8192,0.07123821973800659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,7168,0.09612177477942573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,7168,0.07525244686338636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,10240,0.11114222473568386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,6144,0.11518311500549316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,8192,0.09162044525146484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,5120,0.09682666593127781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,6144,0.05494755506515503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,5120,0.08181689182917277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,5120,0.04689244429270426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,4096,0.08128977484173246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,5120,0.060363557603624135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,8192,0.15549511379665798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,3584,0.06897333595487806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,4096,0.06505955590142144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,51200,0.4976035753885905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,6144,0.09613600042131211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,7168,0.08005866739484999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,4096,0.05026311013433668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,3584,0.05172622203826904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,3584,0.03529777791765001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,2560,0.05077155431111654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,3584,0.046032889021767505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,3072,0.040129777457978986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,3072,0.05024800035688612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,2048,0.04174222217665779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,2560,0.027246221899986267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,3072,0.05993600024117363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,2048,0.03563022282388475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,1536,0.03224533465173509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,2048,0.023119111855824787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,6144,0.0699582232369317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,3072,0.03155288762516446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,1024,0.023639111055268183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,1536,0.019655999210145738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,1536,0.026756443911128577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,1536,0.025954667064878676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,2560,0.03811644514401754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,4096,0.03944799966282315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,1024,0.02014755540423923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,1024,0.02065955599149068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,512,0.01350400017367469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,768,0.013460444079505073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,768,0.015848888291252982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,768,0.017633777525689866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,256,0.008274666965007782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,512,0.012874667015340594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,512,0.01274311128589842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,512,0.014469333820872836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,128,0.006603555546866522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,256,0.010784000158309937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,256,0.010060444474220276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,256,0.012439111040698158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,1024,0.015258666541841296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,2560,0.03538044293721517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,128,0.008160889148712158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,128,0.01069777790043089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,64,0.010037333601050908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,32,0.010049777726332346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,2048,0.03103733393881056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,768,0.018943111101786297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,128,0.010737777584128909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,32,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,65536,1.0872942606608074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,65536,0.5146089129977757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,16384,0.28025510576036244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,51200,0.3909182283613417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,51200,0.8567182223002116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,65536,0.43498934639824766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,64,0.005921777751710679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,16384,0.1307342184914483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,12288,0.2095635599560208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,16384,0.10905955235163371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,12288,0.09798044628567165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,51200,0.33739201227823895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,12288,0.08541511164771186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,10240,0.17547289530436197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,10240,0.10202666785981919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,8192,0.14183021916283503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,16384,0.13652533955044216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,10240,0.08848533365461562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,7168,0.12386755148569743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,12288,0.10287200080023871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,10240,0.08821688758002387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,8192,0.05668355358971489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,7168,0.07846044169531928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,7168,0.05161955621507433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,6144,0.105877333217197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,7168,0.06515466504626803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,6144,0.07006400161319308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,5120,0.0897315541903178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,8192,0.0894355575243632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,6144,0.05736533138487074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,5120,0.06402489211824205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,8192,0.0740106635623508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,4096,0.07299111286799113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,5120,0.05054400033420987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,51200,0.3901066780090332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,4096,0.052595555782318115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,6144,0.046559111939536206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,65536,0.5090364350212945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,3584,0.04183377822240194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,3584,0.029489778810077246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,5120,0.03940622342957391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,3072,0.04109422365824381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,3072,0.02647999922434489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,3072,0.03487822082307603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,4096,0.03303022185961405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,4096,0.04262577825122409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,2560,0.03190666768285964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,2048,0.03781333234575059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,3072,0.05648889144261678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,2048,0.028593778610229492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,2048,0.027127110295825537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,2560,0.04699822266896566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,3584,0.06482311089833577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,1536,0.0218231115076277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,1024,0.021320889393488567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,2560,0.028142223755518597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,1536,0.022106667359670002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,1536,0.016561778055297006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,2560,0.030964444080988567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,3584,0.03875733415285746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,2048,0.01958488921324412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,768,0.013891556196742587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,512,0.011782222323947482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,768,0.015594666202863058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,512,0.011060444845093621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,512,0.013223999904261695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,512,0.011345778074529437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,1024,0.01685244507259793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,1024,0.017551110850440133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,1536,0.029849777619043987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,128,0.0041599998043643106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,256,0.009586667021115622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,256,0.010758221977286868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,128,0.009739555418491364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,64,0.0035182221068276297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,128,0.010043555663691627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,32,0.003546666767862108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,1024,0.013407111167907715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,64,0.00906044410334693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,32,0.008729777402347988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,256,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,65536,0.3229813310835096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,768,0.012085333466529846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,65536,0.17722133795420328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,51200,0.25153689914279515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,128,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,65536,0.15596800380282932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,16384,0.09089689122305976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,256,0.007280000381999546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,16384,0.04754399922158983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,51200,0.12699377536773682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,16384,0.042903999487559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,16384,0.05146844519509209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,12288,0.036152889331181846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,12288,0.0335440006521013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,10240,0.05315999852286445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,12288,0.040525333748923406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,51200,0.13707110616895887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,8192,0.044904887676239014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,768,0.016171556380059984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,10240,0.03679022192955017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,8192,0.03596533338228861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,8192,0.024246222443050806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,8192,0.03228622343805101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,7168,0.031587554348839655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,65536,0.1714471048778958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,51200,0.13111021783616808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,6144,0.033955554167429604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,7168,0.02676888969209459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,10240,0.029370665550231934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,5120,0.028328001499176025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,6144,0.030565334690941706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,12288,0.0711031092537774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,7168,0.03831377625465393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,4096,0.024542222420374554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,5120,0.0233608881632487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,4096,0.021563554803530376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,7168,0.028459555572933618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,4096,0.020267556111017864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,4096,0.015825778245925903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,6144,0.019711999429596793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,6144,0.025166221790843542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,3584,0.014462222655614218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,3584,0.018887110882335238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,5120,0.025049777494536504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,3072,0.017701332767804463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,3072,0.01718133356836107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,2560,0.015812445018026564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,3072,0.013659555878904132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,2560,0.015775110986497667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,2560,0.015929778416951496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,2560,0.012520888613329994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,2048,0.013146667016877068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,2048,0.01368622233470281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,2048,0.014559111661381192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,3584,0.021310221817758348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,10240,0.03983911209636264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,3584,0.018366222580273945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,1536,0.013068444199032254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,3072,0.018583110637134977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,1024,0.00831200016869439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,1024,0.009376889301670922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,1024,0.011388444238238864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,768,0.006937777830494775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,1536,0.010473777850468954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,768,0.007683555285135905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,512,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,768,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,5120,0.017878222796652053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,1536,0.009512888888518015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,256,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,512,0.006738666858938005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,512,0.00902133352226681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,256,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,256,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,1024,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,768,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,128,0.009001777403884465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,64,0.0028764444092909494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,128,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,32,0.0028862222615215513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,512,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,64,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,65536,0.2538062201605903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,2048,0.011215999722480774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,1536,0.01144088887506061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,65536,0.14668088489108616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,65536,0.1339448822869195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,51200,0.19966667228274873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,51200,0.11241332689921062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,16384,0.06690222024917603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,32,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,51200,0.1097537808948093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,256,0.00906311141120063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,12288,0.052206221554014415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,16384,0.04156000084347195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,16384,0.037772443559434675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,16384,0.041469332244661115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,51200,0.10336799754036798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,12288,0.029811554484897192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,12288,0.032724443409177996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,10240,0.036772443188561335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,10240,0.02574666672282749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,128,0.0032008888406885993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,10240,0.029330667522218492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,8192,0.021912000245518152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,8192,0.03077510992685954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,7168,0.0328435566690233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,8192,0.025557332568698462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,12288,0.03193688723776076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,6144,0.029098666376537744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,7168,0.023522666758961145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,6144,0.025260445144441392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,65536,0.13556978437635633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,128,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,6144,0.021494223011864558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,10240,0.044844445255067616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,7168,0.02184355590078566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,4096,0.02056711084312863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,5120,0.016371554798550077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,8192,0.03744533326890733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,4096,0.019467555814319186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,3584,0.01758311192194621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,4096,0.01479822231663598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,6144,0.018361777067184448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,3584,0.01611911091539595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,3584,0.013599110974205865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,7168,0.027327999472618103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,5120,0.01936177743805779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,3072,0.016015999846988253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,3072,0.012150222228633033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,3072,0.015551111764378019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,4096,0.017084444562594097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,5120,0.025448888540267944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,2560,0.013626666532622443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,2048,0.011337777806652917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,3072,0.015450666348139444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,5120,0.02203111184967889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,1536,0.009716444545321995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,2048,0.009992889232105678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,2048,0.013744889034165276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,2560,0.013192888763215808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,1536,0.008671110702885522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,1024,0.007582222421964009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,1536,0.01240799989965227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,2560,0.011500444677140979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,2560,0.014192000031471252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,1024,0.010370666782061258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,3584,0.016167110866970487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,2048,0.012662221988042196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,768,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,512,0.004554666578769684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,768,0.00740799970097012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,1536,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,512,0.007136888802051544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,512,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,256,0.0037866665257347953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,512,0.00905777762333552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,1024,0.008174222376611497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,128,0.003212444484233856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,1024,0.007638221813572778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,256,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,256,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,64,0.0028728888266616394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,128,0.006697777658700943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,32,0.003216888962520493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,768,0.009351111120647853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,128,0.00850133349498113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,64,0.005622222191757626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,65536,0.21817155679066977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,32,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,65536,0.1395324468612671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,768,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,256,0.006671111202902264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,65536,0.1263173288769192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,16384,0.05989333656099108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,128,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,51200,0.10752533541785346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,51200,0.10433866580327351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,16384,0.039651556147469416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,65536,0.11658577124277751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,16384,0.03774844606717428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,10240,0.038344889879226685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,12288,0.03049422303835551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,51200,0.09034844239552815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,12288,0.02978133492999607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,8192,0.03253511256641812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,10240,0.024653333756658766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,10240,0.031256887647840716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,10240,0.027024888330035742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,7168,0.029174221886528864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,12288,0.04647289050949944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,8192,0.02401422295305464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,7168,0.021348444951905146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,6144,0.026346666945351496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,7168,0.021199999584092036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,6144,0.020592000749376085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,6144,0.01959022217326694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,8192,0.021545777718226116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,16384,0.03643911083539327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,7168,0.02362666692998674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,12288,0.02850133180618286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,5120,0.017605332864655387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,5120,0.015790222419632804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,4096,0.016007110476493835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,4096,0.01630311045381758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,3584,0.015864888827006023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,5120,0.018214222457673814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,3584,0.014345778359307183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,51200,0.173126220703125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,4096,0.018346667289733887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,3584,0.013077333569526672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,3584,0.015248000621795654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,3072,0.013454222016864352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,3072,0.012091555529170565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,6144,0.017487999465730455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,3072,0.014851555228233337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,4096,0.014144889182514615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,2048,0.010699555277824402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,2560,0.010767999622556897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,2560,0.012311110893885294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,2560,0.013691555294725629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,1536,0.008946666287051307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,3072,0.014290667242474027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,2048,0.013239999612172445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,1536,0.009706666900051964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,1536,0.008355555435021719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,8192,0.027422222826215956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,1024,0.007832888927724626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,1024,0.009734222458468543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,1024,0.007680888805124495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,768,0.005696889013051987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,768,0.007019555403126611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,768,0.009436444275909001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,512,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,768,0.007122666471534305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,512,0.006269333263238271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,512,0.00906755526860555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,256,0.003891555385457145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,256,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,1024,0.007624000310897827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,2560,0.01221777747074763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,256,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,128,0.0031999999450312722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,256,0.009018667042255402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,2048,0.00941422250535753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,128,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,64,0.0029120000286234748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,128,0.009097778134875828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,32,0.002879111096262932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,64,0.005631999837027655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,5120,0.02272533377011617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,32,0.005709333138333426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,65536,0.17501778072781035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,65536,0.09001777569452922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,512,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,65536,0.07204088899824354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,51200,0.13038488229115805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,128,0.006839111033413146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,16384,0.045540445380740695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,51200,0.06888800197177462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,51200,0.057504885726504855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,1536,0.012164444559150271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,12288,0.03254400028122796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,16384,0.027722666660944622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,16384,0.022397332721286353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,10240,0.027757333384619817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,12288,0.01924888955222236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,12288,0.0273626661962933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,2048,0.011052444577217102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,8192,0.02516088883082072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,10240,0.023760888311598036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,10240,0.016576000385814242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,10240,0.024420445164044697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,7168,0.021914665897687275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,8192,0.020662221643659804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,65536,0.10068533155653213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,8192,0.02164711058139801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,6144,0.018921777606010437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,7168,0.015205333630243937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,16384,0.03316444489690993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,7168,0.020112888680564035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,12288,0.02128622266981337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,5120,0.016550223032633465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,6144,0.012631110846996307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,6144,0.016708445217874315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,4096,0.014838222000333997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,6144,0.01899466746383243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,5120,0.014924445086055331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,5120,0.016906667086813185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,8192,0.014741332994567024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,4096,0.010750222537252637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,4096,0.01275288893116845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,4096,0.015999999311235216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,3584,0.009712888962692684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,7168,0.018031999468803406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,51200,0.08084533611933391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,3072,0.009032888544930352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,2560,0.00998933364947637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,5120,0.011699555648697747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,3584,0.012260444462299347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,2560,0.010047111246320937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,2048,0.008757332960764566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,2560,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,3072,0.011167111496130625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,3584,0.011744000017642975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,1536,0.007641777396202087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,3584,0.01497066683239407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,2048,0.00795111142926746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,2048,0.012654222548007965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,3072,0.011056889262464313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,1024,0.005576888720194499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,3072,0.014196443888876172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,1536,0.010891555911964841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,768,0.004328888737493091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,1024,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,1024,0.009695111049546136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,2560,0.013644444445768992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,768,0.0064080001579390625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,768,0.008718222379684448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,2048,0.009358222285906473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,512,0.0057591112951437635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,512,0.0063279999627007385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,1536,0.007246221933099959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,1536,0.007667555577225155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,512,0.008366222182909647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,128,0.0032017777363459268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,1024,0.006671111202902264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,256,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,256,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,64,0.0028657778683635923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,128,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,128,0.004948444250557157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,32,0.002966222249799304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,512,0.003871110992299186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,65536,0.150163557794359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,32,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,65536,0.08323288626141019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,51200,0.11687200599246556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,256,0.009008888569143083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,65536,0.09964355495240952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,51200,0.06272533204820421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,128,0.00847111145655314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,64,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,16384,0.041681776444117226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,51200,0.053731554084353976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,256,0.003208000005947219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,16384,0.021811554829279583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,12288,0.03310222095913357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,65536,0.06610488891601562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,16384,0.03271466824743483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,10240,0.028274665276209515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,12288,0.0273671117093828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,768,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,10240,0.022801778382725183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,8192,0.021406221720907424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,10240,0.024263999528355066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,16384,0.025910221868091162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,8192,0.019918221566412184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,7168,0.020595555504163105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,8192,0.014254222313563028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,12288,0.017831110292010836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,7168,0.017687110437287223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,6144,0.017883555756674874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,10240,0.01588355501492818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,6144,0.012181333369678922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,5120,0.015549333559142219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,6144,0.018458666072951425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,51200,0.08022222254011366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,8192,0.021511110994550917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,5120,0.014122666584120857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,4096,0.013526221944226159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,5120,0.011158222125636207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,7168,0.013108444710572561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,4096,0.00978133330742518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,7168,0.019832000136375427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,4096,0.015428443749745687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,6144,0.01554666625128852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,3584,0.01143200033240848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,3584,0.015040889382362366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,3072,0.010900444454616971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,5120,0.016758221719000075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,3072,0.014487110906177096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,3072,0.008362666600280339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,3584,0.01185066666867998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,2560,0.007666666474607255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,2048,0.008099555969238281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,2560,0.013370667066839008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,3072,0.010862222148312463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,2048,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,2048,0.008681777450773451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,1536,0.007285333342022366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,12288,0.019825778073734708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,2560,0.009595555563767752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,1536,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,1536,0.007256000406212277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,4096,0.012739555703269111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,1536,0.01050222251150343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,768,0.004190222256713443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,1024,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,1024,0.006333333336644703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,1024,0.009381333159075843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,512,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,768,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,2048,0.012803555362754397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,1024,0.005292444593376583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,512,0.008743999732865227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,256,0.0032586666444937387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,2560,0.009697777529557547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,3584,0.00923466682434082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,256,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,128,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,256,0.00832444429397583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,768,0.006152888966931238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,64,0.0029324444217814338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,128,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,128,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,512,0.005568888866239124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,64,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,65536,0.12315555413564046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,32,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,65536,0.07577511337068346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,65536,0.06380799743864271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,51200,0.09707822402318318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,768,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,51200,0.05671111080381605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,51200,0.05164711011780632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,128,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,512,0.00600000015563435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,51200,0.07986488607194689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,16384,0.023800000548362732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,12288,0.028927998410330877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,16384,0.021631111701329548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,12288,0.017608000172509086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,256,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,10240,0.025083555115593806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,12288,0.026983999543719824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,32,0.002867555452717675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,10240,0.015441776977645027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,8192,0.02184355590078566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,10240,0.024127110838890076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,8192,0.018775999546051025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,16384,0.03274310959710015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,7168,0.019036443697081674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,12288,0.018319111731317308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,7168,0.016513778103722464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,65536,0.09911733203464085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,6144,0.017130666308932833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,7168,0.012913777596420713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,8192,0.013549333645237817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,8192,0.020975111259354487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,6144,0.01811200049188402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,5120,0.015085332923465304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,6144,0.011784888803958893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,5120,0.010728889041476779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,5120,0.013334222137928009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,4096,0.012060444388124677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,5120,0.016561778055297006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,16384,0.03571822245915731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,7168,0.019430221782790292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,3584,0.010447111394670276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,4096,0.011854222251309289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,4096,0.01531733406914605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,3072,0.010040889183680216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,3584,0.009008000294367472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,3584,0.010846222440401712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,2560,0.008950221869680617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,3584,0.014641778336630927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,3072,0.010727110836240979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,3072,0.013713777893119387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,2048,0.007672888537247975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,2560,0.007641777396202087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,2560,0.01294133315483729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,10240,0.020920889245139226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,4096,0.009380444056457942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,2048,0.011462222370836469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,6144,0.014808000789748298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,1536,0.006984000404675801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,1024,0.004907555464241239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,1536,0.010482666393121084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,2560,0.00868800034125646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,1024,0.006292444550328785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,768,0.00423200014564726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,1024,0.0064426664676931165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,1536,0.0068417779273457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,768,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,512,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,768,0.009015999734401703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,3072,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,512,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,256,0.0032017777363459268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,512,0.00851822230550978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,1024,0.009404444032245213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,2048,0.0074622225430276655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,128,0.0028524444335036804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,256,0.005973333285914526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,768,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,128,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,128,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,128,0.00868622213602066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,512,0.0062977779242727495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,32,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,64,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,65536,0.10624177588356866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,32,0.005281777845488654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,2048,0.008808888494968414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,256,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,65536,0.06723466846677992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,65536,0.06146044201321072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,1536,0.006919999917348226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,64,0.002874666617976295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,65536,0.0990613301595052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,51200,0.07988178067737155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,16384,0.021306667062971327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,12288,0.02612977723280589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,16384,0.032413333654403687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,256,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,51200,0.08505511283874512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,12288,0.016901334126790363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,12288,0.026628444592158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,51200,0.05133600036303202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,16384,0.03211999932924906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,8192,0.019744889603720773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,10240,0.015029332704014249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,10240,0.024078221784697637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,8192,0.017872000734011333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,8192,0.01316177762216992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,8192,0.020968889196713764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,10240,0.02258311046494378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,7168,0.01555199921131134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,7168,0.012439111040698158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,10240,0.019931554794311523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,6144,0.014121777481502958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,16384,0.02032533288002014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,6144,0.017902221944597032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,5120,0.013759999639458127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,12288,0.016527111331621807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,5120,0.012757333616415659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,4096,0.011596444580290051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,5120,0.016535111599498324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,4096,0.011743110915025076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,4096,0.0148062225845125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,4096,0.008774222599135505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,6144,0.01148888882663515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,3584,0.010070222119490305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,5120,0.009934222532643212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,3584,0.00833155545923445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,3072,0.009507555928495195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,7168,0.01756977703836229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,3072,0.009404444032245213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,3072,0.007794666621420119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,2560,0.008605333666006723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,6144,0.015936000479592215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,2560,0.008003555238246918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,2560,0.012078222301271228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,3584,0.010429333481523724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,2560,0.007473777565691207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,2048,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,3072,0.01316088851955202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,51200,0.049511998891830444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,2048,0.011092444260915121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,3584,0.014500444134076437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,1536,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,1536,0.006671111202902264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,1536,0.010030222435792288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,1024,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,768,0.004063111212518481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,1024,0.009358222285906473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,7168,0.01904711127281189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,768,0.005913777897755305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,512,0.003559111307064692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,768,0.009024000002278222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,1024,0.004886222382386525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,512,0.005618666609128316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,512,0.00599911105301645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,512,0.008721777962313758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,2048,0.007717333734035492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,256,0.005919111271699269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,128,0.0029066666546795103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,1536,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,2048,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,768,0.006256888724035687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,128,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,128,0.004976888911591636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,32,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,256,0.003212444484233856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,64,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,1024,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,32,0.005347555710209741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,256,0.0052764444715446895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,256,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,65536,0.0610355536142985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,65536,0.059028446674346924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,65536,0.09897777769300674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,64,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,16384,0.02772355576356252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,128,0.008044444852405125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,51200,0.07902489105860393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,12288,0.02196888956758711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,16384,0.018907555275493197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,65536,0.08590844604704116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,16384,0.03204710947142707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,10240,0.018967999352349173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,12288,0.016191111670600045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,12288,0.015299555328157214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,51200,0.06745511293411255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,8192,0.01719199948840671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,10240,0.014551111393504672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,10240,0.02334311107794444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,51200,0.04770400126775106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,7168,0.013820444544156393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,8192,0.016160888804329764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,8192,0.012877333495352002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,8192,0.020597333709398907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,6144,0.012454222473833295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,7168,0.011755555868148804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,7168,0.014839111102951897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,7168,0.018952889574898612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,5120,0.010789333118332757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,6144,0.013791110780504016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,12288,0.02609866691960229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,6144,0.01788444485929277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,10240,0.017671111557218764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,5120,0.00904088881280687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,5120,0.012111110819710625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,51200,0.046606222788492836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,5120,0.01623288955953386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,3584,0.008999110923873054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,4096,0.00868088917599784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,4096,0.014473777678277759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,3584,0.008022222254011366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,3072,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,3584,0.014064888159434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,6144,0.010543111297819348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,3072,0.00902133352226681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,2560,0.007330666813585494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,3072,0.012656889028019376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,4096,0.009696000152164036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,2048,0.006670222100284364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,2560,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,2560,0.007704888780911763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,2560,0.011736888852384357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,2048,0.007130666739410824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,16384,0.019903999235894945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,2048,0.010741333166758219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,1536,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,1536,0.010065777434243096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,1024,0.003955555458863576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,4096,0.01056177748574151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,1024,0.006068444500366847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,2048,0.007352888584136963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,768,0.0036071112586392295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,1024,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,1024,0.009352888498041365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,3072,0.007634667058785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,768,0.006224888894293044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,768,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,1536,0.006687110910813014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,512,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,256,0.0032355555643637977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,512,0.008617777791288164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,1536,0.00490133340160052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,3584,0.009173333644866943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,128,0.0028408887899584244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,256,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,256,0.00798044436507755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,768,0.005611555443869696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,512,0.0035546666218174826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,128,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,32,0.002568888788421949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,65536,0.07820800277921888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,64,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,32,0.0052675555149714155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,65536,0.05701155794991387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,256,0.004947555561860402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,65536,0.05844799677530924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,51200,0.06058222055435181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,64,0.002536888958679305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,16384,0.023696889479955036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,128,0.008336000144481659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,51200,0.07927022377649943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,512,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,12288,0.019339554839664035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,16384,0.018577777677112155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,16384,0.01940266622437371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,16384,0.0317191117339664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,10240,0.016894222961531747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,12288,0.014885332849290637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,65536,0.09864266713460286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,128,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,8192,0.015238222148683337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,10240,0.023392889234754775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,8192,0.01550044450494978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,8192,0.020337777005301583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,7168,0.013441777891582914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,8192,0.01253600004646513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,12288,0.016193777322769165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,12288,0.026017778449588354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,7168,0.011086222198274402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,7168,0.013506666653686099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,10240,0.01789422167672051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,7168,0.018793778287039865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,51200,0.04690222276581658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,6144,0.010055999788973067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,6144,0.01313066648112403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,6144,0.017542221479945712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,4096,0.01000533335738712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,5120,0.009131555755933126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,5120,0.01622222198380364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,4096,0.010034666293197209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,4096,0.008448888858159384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,3584,0.008477333519193862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,6144,0.011781333221329583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,10240,0.014463999205165438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,3072,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,3584,0.008711111214425828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,3584,0.013752889302041797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,3072,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,2560,0.0069466663731469055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,2560,0.0075502221783002215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,51200,0.044928001032935254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,2560,0.011397333608733283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,2560,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,5120,0.011100444528791638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,3072,0.012131555212868584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,2048,0.00665866666369968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,1536,0.0047404443224271136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,2048,0.010909332997269101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,1536,0.006292444550328785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,3584,0.008040000167157915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,1024,0.003896000070704354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,1536,0.010041777458455827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,2048,0.005703111075692707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,3072,0.00775733341773351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,1024,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,1024,0.005888888819350137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,1024,0.009407999614874521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,2048,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,512,0.0032026666320032547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,4096,0.015098666151364645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,768,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,5120,0.0103511114915212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,1536,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,256,0.0028924445311228433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,768,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,256,0.00518488883972168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,768,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,768,0.005585777676767773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,128,0.0028408887899584244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,512,0.006098666538794835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,128,0.004931555440028508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,64,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,128,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,32,0.002508444504605399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,64,0.005021333280536863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,65536,0.06262133518854777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,256,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,256,0.008344000412358178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,65536,0.05479910969734192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,512,0.005622222191757626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,51200,0.05077066686418322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,512,0.008676444490750631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,128,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,65536,0.09849066866768731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,16384,0.02065333392884996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,51200,0.04320799973275927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,51200,0.04617866542604235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,12288,0.017331555485725403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,16384,0.01921866668595208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,16384,0.031874666611353554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,10240,0.016513778103722464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,12288,0.014470222923490735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,12288,0.025897777742809717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,65536,0.057308442062801786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,8192,0.01479644411140018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,10240,0.01428533262676663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,10240,0.023560000790490046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,8192,0.013720889058378009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,8192,0.01180355581972334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,7168,0.012432888978057437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,16384,0.017649778061442904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,7168,0.010610666539933948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,7168,0.012585777375433179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,7168,0.018558222386572096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,10240,0.015949333707491558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,6144,0.012055111428101858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,6144,0.010055111514197456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,5120,0.010007999837398529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,51200,0.07913066943486531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,8192,0.02036888897418976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,4096,0.009089777866999308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,5120,0.009019555317031013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,5120,0.016226665841208566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,6144,0.011378666592968835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,3584,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,4096,0.00906133320596483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,4096,0.008448888858159384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,3584,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,3584,0.007619555625650618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,3072,0.007932444413503012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,6144,0.01756800048881107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,32,0.004972444640265571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,2560,0.006712889091836081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,3072,0.007721777591440413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,3072,0.012481778032249875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,12288,0.016062221593326993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,2048,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,2560,0.007337777978844113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,2560,0.011413333316644033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,2048,0.006492444624503453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,1536,0.004558222161398994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,2048,0.010776888993051318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,3072,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,1536,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,1024,0.0038728887836138406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,2560,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,4096,0.014190221826235453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,1024,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,768,0.0038684445122877755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,3584,0.013036444783210754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,5120,0.010277333358923594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,1536,0.006724444528420766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,512,0.003220444545149803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,768,0.006188444379303191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,1024,0.00563733321097162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,512,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,256,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,1024,0.009371555513805812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,256,0.005030222237110138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,768,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,128,0.0028497777465316984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,128,0.004954666727119022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,2048,0.0069822221994400024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,128,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,128,0.008375999828179678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,256,0.00850844466023975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,64,0.004951999833186467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,32,0.004912000149488449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,512,0.008678221868144141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,64,0.0026613333159022858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,65536,0.05275200141800774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,1536,0.01037688884470198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,32,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,65536,0.05622488922542996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,51200,0.045069333579805165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,65536,0.0982088910208808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,51200,0.04166488846143087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,16384,0.019015999303923715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,51200,0.07913333177566528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,16384,0.017183999220530193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,768,0.009104889300134447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,16384,0.03137333194414774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,12288,0.015607999430762397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,16384,0.018844443890783522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,12288,0.01549600064754486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,12288,0.013834666874673633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,10240,0.014485332700941296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,51200,0.04526488979657491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,10240,0.012783110969596438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,10240,0.014153778553009033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,10240,0.023019555542204116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,256,0.0058417779703934984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,8192,0.012143111063374413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,8192,0.01182666669289271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,8192,0.020466667082574632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,7168,0.01072533345884747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,512,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,6144,0.01070844464831882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,7168,0.018938667244381376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,8192,0.012847999731699625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,6144,0.009714666340086196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,6144,0.01757066614098019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,7168,0.01221244451072481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,5120,0.00977600034740236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,4096,0.008652444514963362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,5120,0.009381333159075843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,7168,0.011436444189813403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,65536,0.05737511316935221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,3584,0.007974222302436829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,4096,0.008037333687146505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,4096,0.00832622249921163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,4096,0.01367555558681488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,3072,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,5120,0.010052444206343757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,3584,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,3584,0.01313866674900055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,2560,0.005981333139869902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,3072,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,3072,0.012099555797047086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,5120,0.01585422290696038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,2048,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,2560,0.006814222368929122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,12288,0.025955556167496577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,2048,0.007320889168315464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,2048,0.006642666541867786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,6144,0.01071733319097095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,2048,0.01073333372672399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,1536,0.00665244460105896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,1536,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,1024,0.00388355553150177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,1536,0.009714666340086196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,2560,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,1024,0.005621333503060871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,768,0.0035386664999855887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,2560,0.011403555671374003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,1024,0.009371555513805812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,512,0.0032382222513357797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,768,0.006234666539563074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,1536,0.004586666822433472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,3584,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,768,0.009033777647548253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,256,0.0031928889867332247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,512,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,512,0.00866844422287411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,256,0.005968889014588461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,256,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,256,0.00831911133395301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,1024,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,64,0.002513777878549364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,128,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,32,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,768,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,64,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,65536,0.05055555701255798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,512,0.0052977779673205475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,32,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,3072,0.007735110819339752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,51200,0.03691555394066705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,65536,0.04990577697753906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,65536,0.05555644300248888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,16384,0.016548444827397663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,51200,0.04488089018397861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,51200,0.040307554933759905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,51200,0.07875200112660725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,16384,0.018590221802393597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,16384,0.016807110773192514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,16384,0.031193776263131037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,12288,0.013816888961527081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,12288,0.01498311095767551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,10240,0.01311288856797748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,12288,0.025853332546022203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,128,0.0028444443725877335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,10240,0.01253600004646513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,10240,0.014344000154071383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,10240,0.022867555419603985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,12288,0.01461155547036065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,8192,0.012245333029164208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,7168,0.011386666860845355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,8192,0.020270221763186984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,7168,0.01036266651418474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,7168,0.018566222654448617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,65536,0.0984133349524604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,7168,0.010265777508417765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,128,0.004942222187916438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,6144,0.009190222455395592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,5120,0.009685333404276106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,8192,0.010915555887752108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,5120,0.008236444658703273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,5120,0.014785778191354541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,6144,0.010379555324713389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,8192,0.012421333127551608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,128,0.005590222362014983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,4096,0.007635555333561367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,3584,0.007713778151406183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,4096,0.0138586668504609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,3584,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,3584,0.007300444775157505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,3072,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,3584,0.013152889079517789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,5120,0.008973333570692275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,4096,0.00832177781396442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,3072,0.007703999678293864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,3072,0.012082666986518435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,6144,0.009742221898502773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,6144,0.016529776983790927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,2560,0.011549333731333414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,2048,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,2048,0.01071911139620675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,2048,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,2560,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,3072,0.006934222247865465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,1536,0.00666844430896971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,1536,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,1024,0.0038746665749284956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,1536,0.010034666293197209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,1024,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,768,0.003536000019974179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,1024,0.009359111388524374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,1536,0.004562666846646203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,768,0.005372444374693765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,512,0.003206222214632564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,768,0.006144000010357962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,4096,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,2048,0.004945777770545748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,512,0.0052897776994440295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,256,0.0031991110493739447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,512,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,1024,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,128,0.0028568889117903183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,256,0.005016889009210798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,256,0.008035555481910706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,768,0.008999110923873054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,128,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,128,0.008119111259778341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,32,0.002576888849337896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,2560,0.006650666809744305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,65536,0.03649955656793382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,512,0.008672888908121321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,32,0.005053333524200651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,256,0.005736888696750005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,51200,0.029179556502236262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,65536,0.048488001028696694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,65536,0.05486577749252319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,65536,0.09845422373877631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,51200,0.04424622323777941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,51200,0.07864444123374091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,2560,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,64,0.004944889081848992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,16384,0.01754666699303521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,16384,0.030943999687830608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,12288,0.013457777599493662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,12288,0.014134221606784396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,128,0.004881777697139316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,16384,0.016398222910033334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,12288,0.013088000317414602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,10240,0.012456888953844706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,8192,0.01035644445154402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,10240,0.023026666707462732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,12288,0.025417778227064345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,16384,0.014184888866212634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,8192,0.010070222119490305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,7168,0.009712888962692684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,8192,0.01994222237004174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,7168,0.010255999863147736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,7168,0.009398221969604492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,6144,0.009713778065310584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,7168,0.017688888642523024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,6144,0.009696889254781935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,5120,0.0087013335691558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,51200,0.039001779423819646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,8192,0.010741333166758219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,5120,0.00813600007030699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,5120,0.008735111190213097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,5120,0.01478844384352366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,4096,0.007832888927724626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,6144,0.008789333204428354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,6144,0.016523554921150208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,3584,0.007076444725195567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,64,0.0028684443483750024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,3584,0.007300444775157505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,10240,0.012764444781674279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,3072,0.006614222294754452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,3584,0.008011555506123437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,4096,0.007648888561460707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,4096,0.01313244468635983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,2560,0.005577777822812398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,3072,0.007637333538797166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,10240,0.011790222591824003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,2048,0.0052764444715446895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,2560,0.007416889071464539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,2560,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,2560,0.011406222151385413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,3584,0.012770666844315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,2048,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,2048,0.006326222171386083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,2048,0.01073155552148819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,1024,0.00388711111413108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,1536,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,1536,0.009720000127951304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,4096,0.008376888930797577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,768,0.0035617777870761026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,1024,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,1024,0.009438222481144799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,768,0.005993777679072485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,512,0.0032462223122517266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,1536,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,3072,0.012456888953844706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,1024,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,512,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,256,0.002973333415057924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,512,0.008655110994974772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,256,0.005702222386995952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,128,0.0028453332682450614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,256,0.008057778080304464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,128,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,128,0.008060444560315874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,128,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,64,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,768,0.005585777676767773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,256,0.005008000052637524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,65536,0.026081777281231348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,32,0.004951111144489712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,3072,0.0069857777820693115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,65536,0.04691377613279554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,51200,0.021597334080272253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,65536,0.054479112227757774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,65536,0.09822933541403876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,51200,0.037703110112084284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,51200,0.04419111212094625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,16384,0.012083555261294046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,768,0.008854221966531541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,16384,0.016217778126398723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,16384,0.01590222285853492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,512,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,64,0.0028817777832349143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,12288,0.012145777543385824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,10240,0.0105262224872907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,12288,0.013471999930010902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,12288,0.025055999557177227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,32,0.002554666664865282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,1536,0.004559111264016894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,8192,0.009379555781682333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,10240,0.02198666665289137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,12288,0.01187111106183794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,8192,0.009717333647939894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,7168,0.010034666293197209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,8192,0.0196435550848643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,7168,0.01037688884470198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,7168,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,6144,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,7168,0.017200888858901132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,6144,0.009380444056457942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,10240,0.012135999898115793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,10240,0.01110755569405026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,6144,0.016229333149062265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,5120,0.009033777647548253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,8192,0.011053333679835001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,16384,0.03091999888420105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,5120,0.014695111248228284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,4096,0.008366222182909647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,4096,0.0076657773719893555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,3584,0.007095110913117726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,4096,0.013518222504191928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,3584,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,6144,0.008767111433876885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,3072,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,3584,0.012806221842765808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,3072,0.007714666426181793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,4096,0.007329777710967594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,5120,0.008053333395057255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,3072,0.012068444656001197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,2560,0.007333333293596904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,2560,0.006609777609507243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,2048,0.006608888920810487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,2560,0.011462222370836469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,51200,0.07861244678497314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,2048,0.006268444574541516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,1536,0.005581333405441708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,2048,0.010802666346232096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,2048,0.006765333314736684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,3072,0.006984000404675801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,1024,0.004607111215591431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,1536,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,1536,0.010053333308961656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,1024,0.005579555614127054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,1024,0.006351999938488007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,1024,0.009180444810125563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,3584,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,768,0.0053093334039052325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,768,0.006171555568774541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,768,0.009018667042255402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,2560,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,1536,0.00655644428398874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,256,0.0032213332338465583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,512,0.006000888844331105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,512,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,512,0.00868977771864997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,128,0.0028880000528362063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,768,0.003926222109132343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,256,0.008252444366614023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,5120,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,128,0.004906666775544484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,64,0.0026906666656335196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,32,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,512,0.0036026665733920205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,768,65536,0.02403555479314592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,64,0.004907555464241239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,32,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,768,51200,0.019644444187482197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,65536,0.05410666598214043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,65536,0.04629155662324694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,65536,0.09775288899739583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,256,0.004933333231343163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,768,16384,0.011396444506115384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,51200,0.03700000047683716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,51200,0.07867288589477539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,128,0.005671999934646819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,768,12288,0.010960000256697336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,128,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,16384,0.030849777989917334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,768,10240,0.009712888962692684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,12288,0.013441777891582914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,12288,0.025311999850802954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,10240,0.01105955574247572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,10240,0.012088889049159156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,768,8192,0.00868977771864997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,10240,0.021651556094487507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,51200,0.043755554490619235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,768,7168,0.009399111072222391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,8192,0.009705777797434065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,16384,0.015871110889646742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,7168,0.010049777726332346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,7168,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,7168,0.01754577789041731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,256,0.005608888963858287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,6144,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,768,5120,0.008038221961922115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,6144,0.009699555734793345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,8192,0.01108444482088089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,16384,0.014659555421935188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,5120,0.00887111077706019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,5120,0.01480888823668162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,768,6144,0.00869422240389718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,4096,0.007536888950400882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,768,3584,0.007009777757856581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,4096,0.013447110851605734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,6144,0.016169778174824186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,3584,0.006989333364698622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,3584,0.00796888851457172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,3584,0.01276177747382058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,8192,0.019091554813914828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,768,4096,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,768,2560,0.00590133335855272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,3072,0.0070053330726093715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,3072,0.012091555529170565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,4096,0.008030222521887885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,768,2048,0.004949333353175057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,2560,0.006526222245560751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,2560,0.01143022212717268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,2048,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,768,3072,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,2048,0.010041777458455827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,5120,0.008064889245563084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,3072,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,768,1024,0.004280888785918553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,2560,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,1024,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,1024,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,768,1536,0.004937777916590373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,768,768,0.0038853333228164246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,1536,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,1536,0.009718221922715506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,768,0.006120000034570694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,768,512,0.0032346666687064697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,512,0.0052871112194326185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,1024,0.009360888765917884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,512,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,768,256,0.003154666680428717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,12288,0.012126222252845764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,768,0.00902133352226681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,256,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,256,0.005607999861240387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,768,128,0.0029004443850782183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,256,0.008040000167157915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,128,0.00565244464410676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,128,0.004954666727119022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,768,64,0.0028826666788922418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,128,0.008014221986134848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,64,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,512,65536,0.017528888252046373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,65536,0.045888887511359326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,2048,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,768,32,0.002570666579736604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,512,51200,0.01604977746804555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,65536,0.09752266936832005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,1536,0.006721777634488211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,32,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,51200,0.03679377834002177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,51200,0.0785822206073337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,16384,0.014488000008794995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,512,12288,0.00979377743270662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,16384,0.01585155559910668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,16384,0.03046755658255683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,12288,0.013464888764752282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,12288,0.012192000117566852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,512,10240,0.00905866672595342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,12288,0.024380443824662104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,512,0.008378666308191087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,512,16384,0.010435555544164447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,10240,0.011899555722872416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,10240,0.01107911103301578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,10240,0.021683555510309007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,512,7168,0.00907199995385276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,8192,0.01071911139620675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,8192,0.010055111514197456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,8192,0.018921777606010437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,512,6144,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,7168,0.010384000009960597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,7168,0.009405333134863112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,7168,0.01754488878779941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,512,5120,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,6144,0.00868888861603207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,768,0.005294222384691238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,6144,0.016543999314308167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,512,4096,0.007338666253619724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,5120,0.00832444429397583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,5120,0.014475555883513557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,4096,0.007344000041484833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,4096,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,512,3584,0.0070017774899800616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,65536,0.0487697786755032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,3584,0.012818666795889536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,512,8192,0.008438222110271454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,5120,0.00869688888390859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,3072,0.0069822221994400024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,4096,0.013453333742088742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,3072,0.007646222081449296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,512,2560,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,6144,0.00941777808798684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,2560,0.006566222343179915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,512,2048,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,3584,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,2048,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,2048,0.010642666783597736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,512,1536,0.004223111189073986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,3584,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,1536,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,2560,0.011420444481902652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,512,3072,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,512,1024,0.003907555507289039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,1536,0.010034666293197209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,1024,0.00563466673096021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,1024,0.006320888797442119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,1024,0.009069333473841349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,2048,0.0069955554273393415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,768,0.006304889089531369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,768,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,768,0.00903733323017756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,2560,0.0073315559162033936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,1536,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,512,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,512,0.005976888868543837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,512,0.008354666332403818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,256,0.005635555419656966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,3072,0.011416888899273343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,256,0.008023111356629265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,512,512,0.003185777821474605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,128,0.004590222405062782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,512,64,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,128,0.00563733321097162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,512,32,0.002519999941190084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,512,256,0.002879111096262932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,64,0.004591111093759537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,256,65536,0.017443555924627516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,32,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,512,768,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,512,128,0.0028568889117903183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,256,51200,0.01479644411140018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,65536,0.046851555506388344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,65536,0.04247555467817518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,65536,0.09686933623419867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,256,16384,0.009709333380063375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,128,0.008040000167157915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,51200,0.07804799742168851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,16384,0.014483556151390076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,16384,0.01588000026014116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,16384,0.030072000291612413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,256,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,12288,0.013394667042626275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,12288,0.012300444145997366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,256,10240,0.007636444436179266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,12288,0.024433778391944036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,51200,0.037392000357309975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,10240,0.0118568887313207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,256,8192,0.006994666324721442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,10240,0.01109777804878023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,10240,0.021557332740889654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,256,7168,0.006727111008432176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,8192,0.010846222440401712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,8192,0.009798222117953831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,8192,0.018587556150224473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,256,6144,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,7168,0.010057777994208867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,7168,0.017872888180944655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,256,5120,0.006727111008432176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,6144,0.009689778089523315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,6144,0.016182222300105624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,256,4096,0.006308444258239534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,5120,0.009019555317031013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,5120,0.008020444048775567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,5120,0.0148044443792767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,256,3584,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,4096,0.007327111230956183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,4096,0.01350844403107961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,3584,0.0069511110583941145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,3584,0.007686222592989604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,3584,0.012098666694429187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,6144,0.00868088917599784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,51200,0.03540088733037313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,3072,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,3072,0.006989333364698622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,3072,0.01145066652033064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,256,12288,0.00798044436507755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,2560,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,256,2048,0.006156444549560547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,2560,0.011400000088744693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,7168,0.009371555513805812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,256,1536,0.0042462220622433555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,2048,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,2048,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,2048,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,256,1024,0.0038702223036024305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,1536,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,256,2560,0.006264888991912206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,1536,0.010064889159467485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,256,768,0.0035626664757728577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,1024,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,1024,0.005625777774386936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,2560,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,4096,0.00832799987660514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,256,512,0.0032115555885765287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,768,0.005305777821275923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,768,0.009027555584907532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,512,0.0059262220230367445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,256,3072,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,512,0.008356444537639618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,1536,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,256,128,0.0028764444092909494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,256,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,128,0.004560889055331548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,768,0.006269333263238271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,256,64,0.002505777817633417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,256,256,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,256,32,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,64,0.004934222333961063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,128,65536,0.01514311134815216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,256,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,32,0.004937777916590373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,65536,0.04620711008707682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,1024,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,65536,0.09640622138977051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,128,0.007836444510353936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,51200,0.03394044438997904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,512,0.0052826665341854095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,128,16384,0.008159110943476358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,51200,0.0767600006527371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,256,0.004975111120276981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,16384,0.015871110889646742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,16384,0.014481777946154276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,16384,0.029861334297392104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,65536,0.04249422086609734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,128,51200,0.012472888661755456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,12288,0.013131555583741931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,128,10240,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,128,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,51200,0.0373253325621287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,128,8192,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,128,12288,0.007620444728268518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,8192,0.009708444277445475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,8192,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,128,7168,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,12288,0.0247022228108512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,7168,0.009355555805895064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,7168,0.017622222503026325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,7168,0.010068444742096795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,128,6144,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,6144,0.008748444418112436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,12288,0.012120889292822944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,6144,0.016362667083740234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,10240,0.011754666765530905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,5120,0.00798044436507755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,128,4096,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,5120,0.014208889669842191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,4096,0.007692444655630324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,128,5120,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,8192,0.019122666782803006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,4096,0.008281778130266402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,4096,0.0131217779384719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,10240,0.011056889262464313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,3584,0.00700711127784517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,3584,0.013099555340078143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,5120,0.008752889103359645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,3072,0.007000000112586551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,3072,0.007687110867765214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,3072,0.011747555600272285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,6144,0.009714666340086196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,2560,0.00674044465025266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,128,2048,0.005628444254398346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,2560,0.011415999796655444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,10240,0.02158311174975501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,3584,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,128,1536,0.004285333471165763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,2048,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,2048,0.006335111127959357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,2048,0.010747555229398938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,128,1024,0.003547555456558863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,128,2560,0.0063324446479479475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,1536,0.010031110710567897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,128,3584,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,1024,0.0063182223174307085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,1024,0.0090506664580769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,2560,0.007287999822033777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,128,3072,0.005965333431959152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,768,0.005990222096443176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,768,0.0053137776752312975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,768,0.00904622260067198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,512,0.005240889059172736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,1536,0.0057502223385704895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,512,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,128,768,0.003576888806290097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,256,0.004593777573770947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,128,128,0.0028773333049482773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,256,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,128,512,0.003228444399105178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,128,0.004600000050332812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,128,64,0.0025173332542181015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,128,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,512,0.005979555762476391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,1536,0.006729777902364731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,64,65536,0.011462222370836469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,64,0.004581333448489507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,64,51200,0.010450666977299584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,1024,0.005718222094906702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,64,16384,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,256,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,65536,0.042187554968727954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,64,12288,0.007305777735180325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,64,10240,0.0069857777820693115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,51200,0.033597333563698664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,64,8192,0.006268444574541516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,16384,0.014476444986131458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,64,7168,0.006670222100284364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,10240,0.011067555182509951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,128,32,0.002874666617976295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,128,256,0.0028657778683635923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,8192,0.009656888743241629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,64,5120,0.00592533333433999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,64,4096,0.00563822231358952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,6144,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,64,3584,0.005680888891220093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,4096,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,64,3072,0.005640000104904175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,3584,0.006858666737874349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,64,2560,0.005668444352017508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,3072,0.00674044465025266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,64,2048,0.005585777676767773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,2560,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,64,1536,0.004563555535342959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,32,0.004936000125275718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,5120,0.007822222179836696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,1536,0.00562755556570159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,1024,0.00555644432703654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,64,768,0.003552000141806073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,12288,0.012082666986518435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,64,512,0.003216000066863166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,64,6144,0.006308444258239534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,64,256,0.0028497777465316984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,768,0.005247999810510212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,64,128,0.0025155554629034465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,512,0.005285333428117964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,64,64,0.002513777878549364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,64,1024,0.003910222401221593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,64,32,0.002556444456179937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,128,0.004596444467703502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,32,65536,0.01110755569405026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,64,0.004972444640265571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,32,51200,0.010443555812040964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,32,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,32,16384,0.0070017774899800616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,65536,0.041746666034062706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,32,12288,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,51200,0.03433155682351854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,32,10240,0.006572444405820634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,2048,0.006268444574541516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,256,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,10240,0.011072888970375061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,32,7168,0.006341333190600078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,8192,0.009692444569534725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,128,0.005649777750174205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,32,6144,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,32,5120,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,7168,0.009376889301670922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,32,4096,0.005640888793600931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,7168,0.009347555538018545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,16384,0.014536889062987434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,5120,0.008023111356629265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,32,3072,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,12288,0.01180266671710544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,4096,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,3584,0.007019555403126611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,3072,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,32,1536,0.004263999975389904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,2560,0.00666933341158761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,32,1024,0.0038773334688610504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,1536,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,6144,0.00867377801073922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,32,3584,0.006262222097979651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,1024,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,32,512,0.003216888962520493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,32,256,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,32,8192,0.006335999816656113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,32,2560,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,32,2048,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,512,0.0052844443255000645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,256,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,32,32,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,2048,0.006310222049554189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,64,0.004955555415815777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,32,0.00480266660451889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,768,0.005249777601824866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,32,64,0.0025351111673646504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,51200,0.9734515084160699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,128,0.004575110971927643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,12288,0.22663466135660806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,32,768,0.003552888830502828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,51200,0.4373146692911784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,16384,0.15824088785383436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,12288,0.12190400229560004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,10240,0.18858222166697183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,16384,0.32434132364061147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,8192,0.15541422367095947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,10240,0.12587466504838732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,16384,0.1630800035264757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,7168,0.13372977574666342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,8192,0.11261867152320014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,8192,0.07179466883341472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,6144,0.11497778362698025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,12,32,128,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,7168,0.06332355737686157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,51200,0.4873991012573242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,6144,0.0550799998972151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,4096,0.0807244446542528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,5120,0.07134489218393962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,10240,0.08682399988174438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,3584,0.06938666767544217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,4096,0.03961688942379422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,4096,0.06293777624766032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,3072,0.06007466713587443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,3584,0.035396443472968206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,7168,0.09551644325256348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,3584,0.052786668141682945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,2560,0.05082044336530897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,3072,0.03604000144534641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,3072,0.04587555593914456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,6144,0.08285511202282377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,2560,0.027502222193611994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,2560,0.040362666050593056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,5120,0.04693866769472758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,2048,0.023223110371165808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,1024,0.02312266661061181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,1536,0.018940443793932598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,1536,0.02404088940885332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,1024,0.019318222999572754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,12288,0.10410044590632121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,5120,0.09669244289398193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,2048,0.04154310954941644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,768,0.014412444498803882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,768,0.013898667362001208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,1536,0.03227377931276957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,256,0.008159110943476358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,2048,0.03496088915401035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,512,0.012432888978057437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,128,0.006608888920810487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,256,0.01071999967098236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,256,0.010216000179449717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,64,0.005751111027267244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,128,0.010671111444632212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,768,0.01889600025282966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,128,0.007975111405054728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,64,0.009366222553782994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,512,0.013501333693663279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,32,0.010057777994208867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,512,0.013148444394270578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,1024,0.015584000282817416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,65536,1.1005333794487846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,65536,0.5133751233418783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,32,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,51200,0.8674942122565376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,51200,0.39030665821499294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,12288,0.21167288886176217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,16384,0.12694311141967773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,16384,0.13036533196767172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,65536,0.4245137903425429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,8192,0.14145867029825845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,10240,0.17561599943372938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,10240,0.08936444256040786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,7168,0.12426843908098008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,8192,0.08915822373496161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,12288,0.09803733560774062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,6144,0.10713777939478557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,7168,0.05141333407825894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,12288,0.08585777547624375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,6144,0.04500000013245476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,6144,0.06672800249523587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,5120,0.09249422285291885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,4096,0.07456711265775892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,5120,0.03892177674505446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,5120,0.05758400095833672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,16384,0.2805599901411268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,4096,0.03307022319899665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,4096,0.05219022101826138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,7168,0.07601244582070245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,3072,0.056722667482164174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,3584,0.0294773346847958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,3584,0.042913777960671314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,2560,0.047107554144329496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,3072,0.02659911082850562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,3072,0.03734044565094842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,2048,0.039183109998703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,2560,0.023079999619060095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,2560,0.03386133246951633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,2048,0.01992266707950168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,8192,0.05778933233684964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,2048,0.0292577776643965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,1024,0.021627555290857952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,1536,0.01681866745154063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,768,0.016492444607946608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,1024,0.01330133361948861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,1024,0.01738755570517646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,512,0.011728888584507836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,768,0.013041777743233575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,10240,0.11949155065748428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,256,0.007296000089910295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,512,0.011239999698268043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,768,0.012429333395428128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,1536,0.03051999873585171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,256,0.009633777870072259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,256,0.008341333104504479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,64,0.0033039999090962936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,3584,0.06401155392328899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,128,0.009727111293209923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,1536,0.020012444920010038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,64,0.008367111285527548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,32,0.008884444832801819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,65536,0.32244356473286945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,51200,0.35840710004170734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,51200,0.25077422459920246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,65536,0.17553955978817412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,128,0.003899555653333664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,16384,0.09108710951275295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,51200,0.13110044267442492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,32,0.003593777616818746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,128,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,16384,0.04796444376309713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,16384,0.04996177885267469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,10240,0.052556445201238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,12288,0.033379554748535156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,12288,0.03638755612903171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,8192,0.04399999976158142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,10240,0.02921066681543986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,10240,0.046041776736577354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,8192,0.024665777881940205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,512,0.010779555473062726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,65536,0.15943733851114908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,7168,0.022328888376553852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,7168,0.03089600139194065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,51200,0.11537778377532959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,6144,0.01998133295112186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,6144,0.026766222384240892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,5120,0.017833777599864535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,4096,0.024696000748210486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,8192,0.035565333233939275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,4096,0.015942222542232938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,3584,0.02127555509408315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,5120,0.02862488892343309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,3584,0.01906844476858775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,3072,0.018567111757066514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,7168,0.037920888927247785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,5120,0.024655110306209985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,2560,0.015827554795477126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,3072,0.013735999663670858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,3072,0.016358221570650738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,2048,0.012855999999576144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,2560,0.012475555141766867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,2560,0.01553244392077128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,1536,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,2048,0.011328000161382886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,2048,0.013931555880440606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,1024,0.008034666379292807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,1536,0.009586667021115622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,768,0.0069360000391801195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,1536,0.010544000400437249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,1024,0.008367111285527548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,512,0.004939555707905027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,768,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,768,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,256,0.0035457776652442086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,512,0.00665244460105896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,4096,0.021056888831986323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,128,0.0029377777957253983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,256,0.006999111009968652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,256,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,64,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,128,0.00701155596309238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,128,0.005495999836259418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,3584,0.014194667339324951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,1024,0.009554666777451834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,32,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,512,0.007532444265153673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,6144,0.033653332127465144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,65536,0.14511556095547146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,12288,0.07097333007388644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,51200,0.19922399520874023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,65536,0.14300444391038683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,32,0.0028684443483750024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,51200,0.11228177282545303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,16384,0.04138133260938857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,64,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,10240,0.044738666878806226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,12288,0.031896889209747314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,10240,0.025791111919615004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,10240,0.038944890101750694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,8192,0.037788444095187716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,16384,0.06715733475155301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,7168,0.032920890384250216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,8192,0.022031999296612207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,16384,0.04403911034266154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,12288,0.05255822340647379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,6144,0.02900444467862447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,7168,0.020280000236299302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,12288,0.02978488802909851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,6144,0.018227555685573153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,6144,0.023915555742051866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,4096,0.020599999361568026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,5120,0.016304888659053378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,5120,0.021242666575643752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,3584,0.017656000124083627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,4096,0.014528888795110913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,4096,0.019155555301242404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,3072,0.01587288909488254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,3584,0.013609777722093793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,3584,0.017249777913093567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,2560,0.013082666529549493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,7168,0.02665866745842828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,65536,0.2542506588829888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,2048,0.011718221836619906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,2560,0.011403555671374003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,2560,0.014338667194048563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,1536,0.009400000174840292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,2048,0.011177777416176267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,2048,0.012150222228633033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,1024,0.007878222399287755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,1536,0.009597332941161262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,51200,0.10507111416922675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,768,0.0064071110553211635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,1024,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,3072,0.012432888978057437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,1024,0.00831200016869439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,512,0.004580444345871608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,3072,0.014613333675596448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,5120,0.02541333271397485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,256,0.0037226668662495087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,768,0.007339555356237624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,512,0.006614222294754452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,128,0.003192000091075897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,256,0.005978666659858491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,1536,0.009354666703277165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,64,0.0028924445311228433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,8192,0.03142133355140686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,32,0.0030177777840031516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,64,0.005619555711746216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,32,0.005995555470387141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,65536,0.21755466196272108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,256,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,51200,0.17296711603800455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,128,0.0069546666410234236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,65536,0.12768533494737414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,128,0.0052675555149714155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,51200,0.10245600011613633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,768,0.007668444679843054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,16384,0.03939555419815911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,12288,0.046401778856913246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,16384,0.04053066505326165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,10240,0.03851022322972616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,12288,0.028728001647525366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,12288,0.03042222062746684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,8192,0.0322017769018809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,65536,0.14172800381978354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,10240,0.03175110949410333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,16384,0.059902224275800914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,8192,0.026356443762779236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,6144,0.025993777645958796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,51200,0.10756444268756443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,7168,0.02275288932853275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,6144,0.017487999465730455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,6144,0.020206222931543987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,5120,0.015660444895426433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,10240,0.02499377727508545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,7168,0.029153777493370905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,4096,0.016341333587964375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,4096,0.014183110660976835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,512,0.007340444458855524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,5120,0.02270488937695821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,3072,0.014470222923490735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,3584,0.014678221609857349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,3072,0.013088888592190213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,4096,0.018597332967652213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,2560,0.012519111235936483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,8192,0.024455999334653217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,2560,0.011732444167137146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,2048,0.010958222051461538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,2560,0.011796444654464722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,7168,0.01928622192806668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,2048,0.009353777600659264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,2048,0.011392888923486074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,1024,0.007676444119877285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,1536,0.008605333666006723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,3584,0.012497777740160624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,768,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,1024,0.0074684446056683855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,3072,0.01313244468635983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,512,0.004583111239804162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,768,0.007366221812036302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,512,0.007167110840479533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,3584,0.01610577768749661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,256,0.005764444255166584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,1024,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,128,0.0033128888656695685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,256,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,128,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,64,0.0032151111712058387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,128,0.005470222069157495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,32,0.00318755561278926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,1536,0.008995555341243744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,64,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,65536,0.17524354987674287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,32,0.005677333308590784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,1536,0.008791111409664154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,5120,0.01771111124091678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,51200,0.13103911611768934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,65536,0.09273599916034275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,65536,0.07066844569312201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,16384,0.04547377758555942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,51200,0.0581208864847819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,12288,0.03278133273124695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,16384,0.02739022175470988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,10240,0.02751911183198293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,12288,0.021495999561415777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,10240,0.023447111248970032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,8192,0.025039111574490864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,768,0.006312888943486744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,51200,0.06925511360168457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,8192,0.020257777637905545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,8192,0.01482399966981676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,256,0.0039022221333450745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,6144,0.01904088921017117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,7168,0.013741333451535968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,7168,0.018231110440360177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,5120,0.016554666890038382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,6144,0.01648266613483429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,5120,0.011786667009194693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,16384,0.0244151105483373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,7168,0.021848888860808477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,4096,0.014402666025691561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,12288,0.01924977699915568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,3584,0.012747555143303342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,4096,0.013159111142158508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,10240,0.0166311115026474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,3584,0.011773332953453064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,5120,0.014686221877733866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,3072,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,2560,0.010032888915803697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,512,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,4096,0.010770666930410596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,2048,0.008551999926567078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,6144,0.012784000072214337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,3072,0.011353777514563667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,2048,0.009342222577995723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,3072,0.010696888797812991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,1536,0.007292444507280986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,1024,0.005592888842026393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,1536,0.007299555672539606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,2560,0.008299555215570662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,1024,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,1024,0.006618666566080517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,512,0.003903999924659729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,2048,0.008352888955010308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,768,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,1536,0.007660444411966536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,256,0.00319377767542998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,512,0.006292444550328785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,512,0.005973333285914526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,256,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,256,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,64,0.002571555475393931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,128,0.005046222358942032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,32,0.002830222249031067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,768,0.004588444613748127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,64,0.005382222019963794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,65536,0.14960177739461264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,3584,0.009744000103738572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,51200,0.11630666255950928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,65536,0.06636622216966417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,128,0.0029404444826973807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,16384,0.04180888997183906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,51200,0.05357422100173103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,51200,0.06265422370698717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,12288,0.03318488929006789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,16384,0.025398222936524287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,12288,0.017887999614079792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,768,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,10240,0.028120001157124836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,12288,0.019847111569510568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,8192,0.0221413340833452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,10240,0.015812445018026564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,10240,0.021827555365032617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,7168,0.020406223005718656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,8192,0.014195554786258273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,128,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,8192,0.0196142229768965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,6144,0.018089777893490262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,7168,0.013061333033773633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,2560,0.01036533299419615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,5120,0.015557333827018738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,6144,0.012254222399658628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,4096,0.013832000394662222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,6144,0.015791111522250705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,32,0.005336888962321811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,5120,0.013836444252067141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,3584,0.012093333734406365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,4096,0.009789333575301701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,3072,0.010690666735172272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,3584,0.009188444250159794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,3584,0.011588444312413534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,2560,0.009694221946928237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,3072,0.010148444109492833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,16384,0.021677333447668288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,2560,0.010032888915803697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,5120,0.011281777587201862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,2048,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,2048,0.00869066682126787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,1536,0.007134222322040134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,1536,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,3072,0.008715555899673039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,7168,0.01663644446267022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,1024,0.006656888872385025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,768,0.004092444562249714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,1024,0.006609777609507243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,768,0.006260444306664997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,4096,0.012688889271683164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,768,0.005976888868543837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,512,0.006330666856633292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,1024,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,256,0.003232888877391815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,256,0.005978666659858491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,2560,0.007720888488822513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,128,0.003168888803985384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,2048,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,64,0.002862222285734283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,128,0.005973333285914526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,32,0.002874666617976295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,128,0.00508977762526936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,65536,0.08186044295628865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,64,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,65536,0.12299111154344346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,32,0.005385777602593104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,512,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,65536,0.0739484429359436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,65536,0.06367644336488512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,256,0.005221333354711533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,51200,0.05165955424308777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,16384,0.021130666136741638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,16384,0.02381688853104909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,12288,0.028740445772806805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,12288,0.01739911072784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,1536,0.006981333096822103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,51200,0.09703022241592407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,16384,0.035720000664393105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,8192,0.021298666795094807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,8192,0.013937777943081327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,8192,0.01850755512714386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,10240,0.02459822263982561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,7168,0.01585422290696038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,6144,0.017165333032608032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,7168,0.012777778009573618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,10240,0.01999200052685208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,5120,0.015374221735530429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,6144,0.01201866649919086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,6144,0.014517333772447375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,4096,0.011731555892361535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,5120,0.01070666644308302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,512,0.0038559999730851916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,5120,0.01313866674900055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,3584,0.010413332945770687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,4096,0.008741333252853816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,3072,0.010381333529949188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,3584,0.008344000412358178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,12288,0.018067555295096505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,2560,0.008911111288600499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,3072,0.009637333452701569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,51200,0.05657688776652018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,2048,0.007801777786678738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,7168,0.019308444526460435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,2048,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,2048,0.007341333561473423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,1536,0.006719111154476802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,3072,0.007968000239796108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,1024,0.005008888741334279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,1536,0.006996444529957241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,10240,0.015494222442309061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,1024,0.006342222293217977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,1024,0.006313777632183499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,768,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,4096,0.012087110843923358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,3584,0.011058666639857821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,1536,0.006959999601046245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,512,0.006309333360857434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,2560,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,768,0.004267555558019214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,128,0.003118222165438864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,256,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,128,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,64,0.0028693332440323303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,128,0.004967111266321606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,32,0.002838222309947014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,64,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,256,0.002955555501911375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,65536,0.10781777567333645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,2560,0.00905777762333552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,51200,0.08569333288404678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,256,0.005305777821275923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,65536,0.06745688782797919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,768,0.006267555471923616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,512,0.003855111284388436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,51200,0.049626668294270836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,12288,0.026130666335423786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,16384,0.020616888999938965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,32,0.005320888840489917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,10240,0.02239022155602773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,12288,0.01644533375898997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,65536,0.061344888475206166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,10240,0.01514488955338796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,51200,0.051186667548285596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,8192,0.01331733332739936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,16384,0.021359999974568684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,7168,0.017249777913093567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,7168,0.012412444584899478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,12288,0.01687111126052009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,6144,0.015842666228612263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,6144,0.011433777709801992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,8192,0.01963911122745938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,512,0.005619555711746216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,8192,0.017303110824690927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,4096,0.011861333416567909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,7168,0.015121777852376303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,4096,0.008650666309727563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,5120,0.013819555441538492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,3584,0.01057866629627016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,3584,0.008157333566082848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,3072,0.009341333475377824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,5120,0.009881777895821465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,3072,0.008664888640244802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,4096,0.011742221812407175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,10240,0.020267556111017864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,2560,0.00830844458606508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,2560,0.007617777420414819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,3584,0.010728889041476779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,2048,0.0075804442167282104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,2048,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,1536,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,16384,0.03192799952295091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,1536,0.006866666591829724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,1024,0.004722666823201709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,2560,0.008606221940782335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,5120,0.012472000386979846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,768,0.004211555752489301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,6144,0.013617777989970313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,768,0.0063164445261160536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,512,0.003576000117593341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,512,0.005977777971161737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,256,0.003216000066863166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,1024,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,2048,0.007679999702506595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,128,0.002883555574549569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,256,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,256,0.0052968888647026485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,128,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,768,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,128,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,64,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,32,0.004957333207130432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,512,0.005614222337802251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,65536,0.05974488788180881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,65536,0.05899288919236925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,51200,0.06782578097449408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,64,0.002569777684079276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,1536,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,16384,0.02777333392037286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,51200,0.046795556942621865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,16384,0.01962933275434706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,16384,0.019111111760139465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,65536,0.08709510829713608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,10240,0.019087110956509907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,12288,0.016221332881185744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,12288,0.01533511115445031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,8192,0.016864889197879367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,10240,0.01458666721979777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,10240,0.019069333871205647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,8192,0.012888000243239932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,3072,0.00776800016562144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,8192,0.01609155535697937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,32,0.0028488888508743713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,6144,0.012444444828563266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,7168,0.011691555380821228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,5120,0.0110880004035102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,7168,0.013808888693650564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,6144,0.013391111459996967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,4096,0.010015111002657149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,1024,0.006400888992680444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,5120,0.012069332930776807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,3584,0.008746667040718926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,4096,0.008362666600280339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,4096,0.011063111325105032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,3072,0.008099555969238281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,7168,0.013759111364682516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,3584,0.008832889298597971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,2560,0.007213333414660559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,3072,0.007713778151406183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,51200,0.04763466782040066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,3072,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,2048,0.006275555739800136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,6144,0.010445333189434476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,1536,0.004940444396601783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,2048,0.0069857777820693115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,2048,0.007297777467303806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,1024,0.0039057777159743835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,1536,0.006605333338181178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,1024,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,3584,0.008031110796663497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,768,0.003567111161020067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,768,0.006311999840868845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,12288,0.021640888518757288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,2560,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,2560,0.007985778152942657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,256,0.0029377777957253983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,512,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,512,0.005323555734422471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,1536,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,128,0.0028568889117903183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,256,0.005221333354711533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,128,0.00591822216908137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,1024,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,32,0.002817777709828483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,64,0.0052728888889153796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,768,0.005608888963858287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,32,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,5120,0.009326222042242685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,51200,0.060193777084350586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,65536,0.05817688835991753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,65536,0.058111998769972056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,16384,0.023639111055268183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,64,0.0026844443960322272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,51200,0.04689866635534498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,128,0.004941333499219683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,12288,0.019263111882739596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,16384,0.019360888335439894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,16384,0.01828266680240631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,10240,0.016671111186345417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,12288,0.014816888504558139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,10240,0.014344000154071383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,10240,0.017096888687875535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,8192,0.014827556080288358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,8192,0.01238222254647149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,7168,0.01368622233470281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,512,0.003240888938307762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,12288,0.01611111064751943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,7168,0.012792000340090858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,65536,0.07787999841901991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,7168,0.011335999601417117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,5120,0.010514666636784872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,6144,0.010075555907355415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,256,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,6144,0.012104888757069906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,4096,0.009321778184837764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,5120,0.009009777671760982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,5120,0.011651555697123209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,51200,0.045026666588253446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,4096,0.008368888662921058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,8192,0.015475556254386902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,3072,0.00797155582242542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,3584,0.00832088871134652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,2560,0.0069439998931354946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,3584,0.008607111043400234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,3072,0.00797333319981893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,2048,0.005655999812814925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,2560,0.007480888730949826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,2048,0.007062222394678328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,1536,0.004600000050332812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,3584,0.008434666527642144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,1024,0.0038977778620190094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,1536,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,3072,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,768,0.003535110917356279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,2560,0.007340444458855524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,768,0.00620888877246115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,768,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,512,0.003244444520937072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,2048,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,256,0.0031973332580592898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,512,0.005998222364319696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,1536,0.006734222173690796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,4096,0.009463110731707679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,256,0.005990222096443176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,256,0.0052426668504873914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,64,0.0028497777465316984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,128,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,1024,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,32,0.0028364445186323593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,128,0.004912000149488449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,65536,0.06370844443639119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,64,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,32,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,65536,0.05725244681040446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,51200,0.0459191103776296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,51200,0.043030222256978355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,128,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,1024,0.006341333190600078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,16384,0.019317333896954853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,12288,0.017255999975734286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,6144,0.011972444752852121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,10240,0.016538666354285348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,12288,0.015839111473825242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,51200,0.0505946675936381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,12288,0.014465777410401238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,8192,0.015125332607163323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,10240,0.015776889191733468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,7168,0.012753778033786349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,512,0.005325333525737126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,8192,0.013918222652541267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,6144,0.011360000405046674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,7168,0.010537777509954242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,7168,0.012771555946932899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,5120,0.010046222143703038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,16384,0.01759377784199185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,6144,0.011126221881972419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,5120,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,5120,0.010040889183680216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,65536,0.05471377902560764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,3584,0.00814844419558843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,4096,0.008389333056079017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,4096,0.008807111117574904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,3072,0.007522666619883643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,3584,0.008103110724025303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,3584,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,2560,0.006385777973466449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,3072,0.007328888608349695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,2560,0.007326222128338284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,2560,0.007296889192528195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,10240,0.014126222994592456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,8192,0.01145066652033064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,2048,0.006679111056857639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,1536,0.004571555389298333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,16384,0.02089866664674547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,3072,0.007690666450394525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,1536,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,4096,0.009343110852771336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,2048,0.005426666802830166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,1024,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,768,0.003540444291300244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,1024,0.006355555521117316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,512,0.0035377778112888336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,768,0.006115555349323485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,768,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,256,0.003185777821474605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,512,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,512,0.005306666509972678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,1536,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,256,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,1024,0.003990222182538775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,64,0.0028782222006056043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,128,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,32,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,128,0.004963555683692296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,65536,0.057738668388790555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,64,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,32,0.00500177757607566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,51200,0.04503644506136576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,65536,0.0527671111954583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,128,0.0029066666546795103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,6144,0.009701333112186855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,256,0.005221333354711533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,51200,0.04523644513554043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,12288,0.01565511193540361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,16384,0.018851555056042142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,16384,0.01716888944307963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,12288,0.015218666858143277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,12288,0.014136888914638095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,10240,0.012783110969596438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,65536,0.05635199944178263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,10240,0.014198222094111972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,7168,0.012102222277058495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,8192,0.011224889092975192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,8192,0.013066666821638743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,6144,0.011081777513027191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,51200,0.04161244299676683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,7168,0.011752000285519494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,5120,0.010299555957317352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,6144,0.009702222214804756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,10240,0.014397333065668741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,6144,0.01034222212102678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,4096,0.008985777695973715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,5120,0.008998221821255153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,2048,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,3584,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,8192,0.012647111382749347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,3072,0.00699733363257514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,4096,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,3584,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,2560,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,3584,0.007350222104125553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,3072,0.0072266666425599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,2560,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,2560,0.00721066693464915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,2048,0.006996444529957241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,2048,0.006309333360857434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,4096,0.008314666648705801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,1536,0.005963555640644497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,1536,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,1024,0.0039013334446483185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,3072,0.007619555625650618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,768,0.003552888830502828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,1024,0.006303999986913469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,16384,0.018989332848125033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,1024,0.005629333357016246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,512,0.0032142222755485107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,768,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,256,0.0032115555885765287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,512,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,1536,0.0043333334227403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,512,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,128,0.002934222213096089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,256,0.005973333285914526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,256,0.004990222139490975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,64,0.002708444371819496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,128,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,128,0.004942222187916438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,65536,0.049759109814961754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,64,0.004950222041871813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,7168,0.01038755559258991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,2048,0.005230222311284807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,51200,0.03706577751371596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,65536,0.05544888973236084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,65536,0.050311999188529126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,768,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,51200,0.044919110006756253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,51200,0.04023022121853299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,16384,0.018248889181349013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,5120,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,10240,0.013390222357379066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,12288,0.013806222213639153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,32,0.0025759999536805684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,12288,0.013792888985739814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,8192,0.012783110969596438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,10240,0.01219911128282547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,10240,0.014138667119873894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,32,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,8192,0.011735111474990845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,16384,0.016747555798954435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,6144,0.010370666782061258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,7168,0.010187555518415239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,16384,0.01685244507259793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,5120,0.009561777942710454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,5120,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,7168,0.011052444577217102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,4096,0.008430221842394935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,5120,0.009022222624884712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,7168,0.010046222143703038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,3584,0.007289778027269576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,6144,0.009640889035330879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,6144,0.009009777671760982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,3072,0.006322666588756774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,3584,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,3584,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,3072,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,3072,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,8192,0.010949333508809408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,2560,0.007367110914654202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,2048,0.007012444237867991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,12288,0.014522666732470194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,4096,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,1536,0.006690666493442323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,1024,0.003948444293604957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,2560,0.005667555663320754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,1024,0.0063386667105886675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,2048,0.005072000126043956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,768,0.0035662220584021676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,768,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,512,0.0031902222997612427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,1536,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,512,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,256,0.002847111059559716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,1024,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,2560,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,128,0.0028560000161329904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,256,0.005616888817813661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,256,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,1536,0.004597333156400257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,128,0.005821333163314396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,32,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,128,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,65536,0.03673511081271701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,512,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,32,0.004910222358173794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,51200,0.029183109601338703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,65536,0.05496355560090807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,65536,0.04862666792339749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,16384,0.014135999812020196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,51200,0.044553776582082115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,51200,0.03865333398183187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,12288,0.01275199982855055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,64,0.002536888958679305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,12288,0.013468444347381592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,10240,0.011128000087208219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,64,0.005048000150256687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,10240,0.012187555432319641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,8192,0.010381333529949188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,4096,0.008340444829728868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,16384,0.016360888878504436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,8192,0.010049777726332346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,7168,0.00963288876745436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,8192,0.0107342220014996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,6144,0.009806222385830348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,7168,0.010321777727868822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,7168,0.00942755573325687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,5120,0.0086666668454806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,6144,0.00941066692272822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,6144,0.009008000294367472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,10240,0.012331555287043253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,4096,0.008042666647169325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,5120,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,4096,0.008345777789751688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,3584,0.006953777538405524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,2048,0.006254222244024277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,3072,0.006267555471923616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,3584,0.008016000191370646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,3584,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,2560,0.005581333405441708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,12288,0.01347644461525811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,3072,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,2048,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,2560,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,2560,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,1536,0.004562666846646203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,2048,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,768,0.005591111050711737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,1536,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,1024,0.003933333274390963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,1024,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,768,0.003536888708670934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,5120,0.009018667042255402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,768,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,1536,0.006642666541867786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,512,0.0032115555885765287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,3072,0.007646222081449296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,1024,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,512,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,256,0.003132444495956103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,512,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,128,0.0028684443483750024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,256,0.005819555371999741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,2048,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,256,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,128,0.005646222167544895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,32,0.0028560000161329904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,128,0.00488800017370118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,4096,0.0075777777367168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,65536,0.026053332620196875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,64,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,51200,0.021506667137145996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,32,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,65536,0.054457777076297335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,16384,0.012104888757069906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,51200,0.04382310973273384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,51200,0.03758755657407973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,12288,0.01167555567291048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,16384,0.016025778320100572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,16384,0.01516355574131012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,64,0.002564444517095884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,12288,0.013280889226330651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,12288,0.012462221913867526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,16384,0.01754844519827101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,10240,0.012073777616024017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,10240,0.011108444796668159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,7168,0.009678222239017487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,8192,0.010715555813577441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,8192,0.009741333623727163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,7168,0.010028444230556488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,7168,0.009345778160625035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,5120,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,768,0.005586666779385672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,6144,0.008882666627566019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,4096,0.007455999652544658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,5120,0.009009777671760982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,5120,0.008346666892369589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,3584,0.006792000184456508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,4096,0.00832177781396442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,4096,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,3072,0.006262222097979651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,3584,0.007273777491516537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,6144,0.00904444439543618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,2560,0.005624889085690181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,3072,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,6144,0.009700444837411245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,2560,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,2560,0.006697777658700943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,8192,0.009354666703277165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,2048,0.006980444822046492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,2048,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,3584,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,1536,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,65536,0.0469359987311893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,1536,0.005998222364319696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,768,0.00423200014564726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,1024,0.006322666588756774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,1024,0.005623111294375525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,512,0.0035377778112888336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,10240,0.010408889088365767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,768,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,256,0.0032177778581778207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,512,0.005591111050711737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,1536,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,256,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,256,0.004945777770545748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,1024,0.004564444637960858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,64,0.0025822222232818604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,128,0.005985777825117111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,32,0.0028355556229750314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,3072,0.007010666860474481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,768,65536,0.02368533280160692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,2048,0.006264888991912206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,64,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,768,51200,0.019239111079110038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,65536,0.05408177773157755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,512,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,128,0.0032177778581778207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,51200,0.04342666599485609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,51200,0.037343998750050865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,768,12288,0.011160888605647616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,16384,0.014767999450365702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,12288,0.013167111410035027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,768,10240,0.009468444519572789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,32,0.00498488876554701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,65536,0.046216888560189136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,768,16384,0.01128888875246048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,10240,0.012117333710193634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,8192,0.010732444624106089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,768,7168,0.00907022257645925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,128,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,7168,0.010031999813185798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,768,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,7168,0.009349333743254343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,12288,0.012098666694429187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,6144,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,768,5120,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,768,8192,0.008731555607583787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,10240,0.011012444065676795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,768,4096,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,5120,0.0086986662613021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,5120,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,4096,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,4096,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,3584,0.007675555845101674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,3584,0.007267555428875818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,6144,0.00868888861603207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,3072,0.006739555547634761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,768,2560,0.005886222339338726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,3072,0.0076248885856734375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,8192,0.009773333039548662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,16384,0.016180444094869826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,2560,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,768,6144,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,768,3072,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,1536,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,1536,0.006344888773229387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,768,1024,0.004250666747490565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,2560,0.007376888559924231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,768,768,0.0038933331767717996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,1024,0.006341333190600078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,2048,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,2048,0.006063999815119638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,768,0.0063546668324205614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,512,0.005964444329341252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,768,256,0.0031884445084465873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,768,3584,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,256,0.005617777920431561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,768,128,0.0029093333416514923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,768,1536,0.004908444566859139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,768,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,768,64,0.00254844439526399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,128,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,768,32,0.0025680000997251934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,512,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,512,65536,0.01759022143152025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,64,0.004969777746333016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,32,0.004931555440028508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,768,2048,0.005241777747869492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,65536,0.04554133282767402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,1024,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,51200,0.03858400053448147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,768,512,0.003485333174467087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,51200,0.036850667662090726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,512,12288,0.010048000348938836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,16384,0.016061334146393668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,128,0.004933333231343163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,16384,0.014460444450378418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,512,10240,0.009019555317031013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,12288,0.013437333206335703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,12288,0.012141333685980903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,10240,0.012079999678664737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,10240,0.010951999988820819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,256,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,512,7168,0.008663111262851292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,8192,0.009707555174827576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,512,6144,0.008325333396593729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,7168,0.010328888893127441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,512,16384,0.010159999959998662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,7168,0.009474666582213508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,512,5120,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,6144,0.00869777798652649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,5120,0.00868000007337994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,5120,0.008007110820876228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,512,8192,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,512,3584,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,4096,0.008340444829728868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,512,51200,0.015987555185953777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,8192,0.01072622256146537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,512,3072,0.006610666712125142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,3584,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,3584,0.006950221955776215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,3072,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,3072,0.0069875551594628235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,6144,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,512,2048,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,2560,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,512,1536,0.004544888933499654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,2048,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,2048,0.006000888844331105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,1536,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,4096,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,65536,0.04979733294910855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,512,768,0.0035546666218174826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,1024,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,1024,0.005616888817813661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,512,512,0.003208000005947219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,768,0.005961777849329843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,768,0.00564533347884814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,2560,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,512,0.005910222315125995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,512,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,512,128,0.0026293332791990707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,256,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,256,0.004596444467703502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,512,1024,0.0038737778862317405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,512,64,0.0028364445186323593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,128,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,512,32,0.002532444480392668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,512,2560,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,64,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,256,65536,0.01754488878779941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,32,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,256,51200,0.014688000082969666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,65536,0.0466302235921224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,512,4096,0.007295110987292395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,256,16384,0.009407999614874521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,51200,0.03740266627735562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,51200,0.034111999803119235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,256,12288,0.008105778031879002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,1536,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,128,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,256,10240,0.0074684446056683855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,12288,0.013456000222100152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,12288,0.012063999970753988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,256,8192,0.007015110717879401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,10240,0.01072622256146537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,512,256,0.0031893334041039148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,256,7168,0.006682666639486949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,65536,0.04502222273084852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,8192,0.00980799976322386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,256,6144,0.007174222005738153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,7168,0.010300444232092964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,256,5120,0.006607999818192587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,16384,0.0158906661801868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,16384,0.014500444134076437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,6144,0.008700444466537898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,5120,0.009030222064918941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,5120,0.008037333687146505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,256,3584,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,4096,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,4096,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,256,3072,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,3584,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,8192,0.01072355525361167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,3584,0.007275555696752336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,256,2560,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,3072,0.007688889073001013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,256,2048,0.005695111221737332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,2560,0.007270221908887227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,6144,0.009339555270142024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,2560,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,2048,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,2048,0.006302222195598815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,1536,0.006572444405820634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,10240,0.012078222301271228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,1536,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,256,768,0.003415999934077263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,1024,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,1024,0.005285333428117964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,768,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,768,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,3072,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,7168,0.009408888717492422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,512,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,256,1536,0.004587555511130227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,512,0.005283555636803309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,256,4096,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,256,0.005643555687533484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,256,64,0.002557333351837264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,256,0.004957333207130432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,128,0.0057608890864584185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,256,32,0.0028453332682450614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,128,0.004935111022657818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,128,65536,0.014930667148696052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,64,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,128,51200,0.01276088919904497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,256,512,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,65536,0.04217777649561564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,128,16384,0.00831200016869439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,256,256,0.0029128889242808023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,51200,0.033956444925732084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,128,12288,0.007746666669845581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,256,128,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,16384,0.014455111490355598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,128,10240,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,256,1024,0.003927111211750242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,12288,0.013105777402718862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,128,8192,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,10240,0.012148444851239523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,10240,0.011074666347768573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,32,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,8192,0.009723555710580613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,51200,0.03736089004410638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,7168,0.009019555317031013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,16384,0.01585422290696038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,6144,0.009728000395827824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,6144,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,128,5120,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,5120,0.008945778012275696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,65536,0.0462337765428755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,7168,0.010024000373151567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,128,6144,0.006274666637182236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,128,3584,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,4096,0.008000888758235509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,12288,0.012131555212868584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,3584,0.0070266665683852295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,128,3072,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,3584,0.007811555431948767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,128,7168,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,128,2560,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,3072,0.007640889121426477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,3072,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,128,2048,0.005620444400442972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,4096,0.007358222372002072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,2560,0.006663110935025745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,128,1536,0.004545777622196409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,2048,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,2048,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,128,1024,0.0038808890514903595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,1536,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,8192,0.010803555448849997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,1024,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,2560,0.007280888656775157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,768,0.0052764444715446895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,128,512,0.0032035555276605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,768,0.005982222242487802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,512,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,1536,0.00684799998998642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,512,0.004951999833186467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,256,0.005619555711746216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,128,4096,0.0058444444504049085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,128,128,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,128,64,0.002512000087234709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,128,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,128,32,0.002817777709828483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,128,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,64,65536,0.01127199994193183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,64,0.004607111215591431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,64,51200,0.010603555374675326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,32,0.004579555657174853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,64,16384,0.006670222100284364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,65536,0.04187822341918945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,64,12288,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,51200,0.03360889024204678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,5120,0.008062221937709385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,256,0.004964444372389051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,64,10240,0.006308444258239534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,16384,0.014139556222491793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,64,8192,0.006683555742104848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,12288,0.012106666962305704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,10240,0.011063111325105032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,64,6144,0.006583999842405319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,64,5120,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,7168,0.009354666703277165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,64,4096,0.005667555663320754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,128,256,0.0028782222006056043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,6144,0.00868444475862715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,5120,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,64,3072,0.00600622221827507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,4096,0.007348444726732042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,1024,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,128,768,0.003559999995761447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,3584,0.006993778049945831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,64,7168,0.006311111152172089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,3072,0.006984000404675801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,2560,0.006640000061856375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,64,1024,0.003548444559176763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,1536,0.005623111294375525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,64,768,0.0035422220826148987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,1024,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,64,512,0.003176888864901331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,768,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,64,2560,0.005994666781690385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,64,2048,0.005642666584915585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,512,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,64,1536,0.0042257776690853965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,64,64,0.0025111111915773815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,256,0.004963555683692296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,8192,0.009716444545321995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,64,32,0.0028391112056043413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,128,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,32,65536,0.011386666860845355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,32,51200,0.010475555227862464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,64,0.004596444467703502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,32,16384,0.006648889018429651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,64,3584,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,65536,0.04212444358401828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,51200,0.033977776765823364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,32,10240,0.007012444237867991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,64,128,0.0026026666164398193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,12288,0.012072888513406118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,32,8192,0.006324444380071428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,32,7168,0.006313777632183499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,10240,0.010764444039927589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,32,6144,0.0063075555695427795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,8192,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,32,5120,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,7168,0.009219555391205682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,32,0.004594666676388847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,6144,0.008692444198661381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,32,12288,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,32,3584,0.005984888722499211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,5120,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,64,256,0.0028497777465316984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,3584,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,32,2560,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,3072,0.0069208890199661255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,32,2048,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,2560,0.006361777583758037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,32,1536,0.004402666870090696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,2048,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,32,1024,0.003645333151022593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,32,3072,0.005624889085690181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,1536,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,1024,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,32,512,0.003212444484233856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,768,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,32,256,0.0028435554769304064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,512,0.005223999834722943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,32,4096,0.005607999861240387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,256,0.004943999979231092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,32,64,0.002529777793420686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,32,768,0.003604444364706675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,32,32,0.002846222163902389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,64,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,2048,0.006241777704821692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,32,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,11,32,128,0.0028693332440323303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,16384,0.32371200455559623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,51200,0.9671822653876411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,16384,0.16304177708095974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,16384,0.13648088773091635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,12288,0.24679467413160536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,12288,0.14275911119249132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,51200,0.4547022183736165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,10240,0.18840000364515516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,12288,0.10335822237862481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,128,0.004961777892377642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,10240,0.14846932888031006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,10240,0.08721599976221721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,7168,0.1331919961505466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,8192,0.07089244657092623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,6144,0.11482132805718316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,7168,0.09261422024832831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,7168,0.06362577941682604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,6144,0.05530133181148105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,51200,0.48867379294501406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,8192,0.15453067090776232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,4096,0.08013421959347196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,5120,0.04665689004792107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,5120,0.06809244553248088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,8192,0.1347608831193712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,3584,0.06892622179455228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,4096,0.039396444956461586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,3072,0.060028447045220264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,3584,0.03647199935383267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,3584,0.05305333269966973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,2560,0.05048266715473599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,5120,0.096669336160024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,16384,0.014152889450391134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,3072,0.04712444543838501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,2048,0.041816890239715576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,2560,0.027450667487250432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,1536,0.03196622265709771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,2048,0.023123555713229712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,2048,0.03427111109097799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,1536,0.02679644525051117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,4096,0.06412088871002197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,1024,0.015241778559154935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,1024,0.019874667127927143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,768,0.0189057770702574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,2560,0.03820711043145921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,512,0.013583111266295115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,768,0.016180444094869826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,512,0.013109332985348172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,1536,0.01919644408755832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,256,0.010046222143703038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,256,0.010416000253624385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,128,0.006595555692911148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,128,0.008020444048775567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,128,0.01072533345884747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,64,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,3072,0.0381351113319397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,32,0.005738666488064661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,768,0.015073777900801765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,32,0.010065777434243096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,512,0.012443555725945367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,6144,0.08552355236477321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,65536,0.5135466787550185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,65536,1.1196355819702148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,64,0.009359111388524374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,16384,0.28132711516486275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,51200,0.3899804486168756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,1024,0.023353776997990076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,16384,0.13014045026567247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,12288,0.20922666125827363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,65536,0.4248390992482503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,12288,0.11520444022284614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,10240,0.1744257741504245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,12288,0.10053866439395481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,16384,0.1129866706000434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,8192,0.14195466041564941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,10240,0.06910400258170234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,8192,0.08887466457155015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,256,0.00814933329820633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,51200,0.847567982143826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,6144,0.10654044151306152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,7168,0.05135022269354927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,7168,0.0749173363049825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,5120,0.09027200274997288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,51200,0.3518853452470567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,10240,0.09967288706037734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,6144,0.06786844465467665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,5120,0.05606044663323296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,7168,0.12388622760772705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,8192,0.0606222218937344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,3584,0.06420622269312541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,4096,0.03295644455485874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,4096,0.049923555718527905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,3072,0.05569689141379463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,3584,0.02904355525970459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,2560,0.047129776742723256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,6144,0.04497155547142029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,3072,0.03803022371398078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,3072,0.02625422179698944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,2048,0.03783288929197524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,2560,0.022870222727457683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,2560,0.03089600139194065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,1536,0.029855112234751385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,2048,0.023895111348893907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,2048,0.02903466754489475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,1536,0.016520889268981088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,1536,0.02233955595228407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,768,0.016178665889634024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,1024,0.01685333251953125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,3584,0.043839110268486865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,512,0.011741333537631564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,768,0.013507555756303998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,768,0.013790222505728403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,256,0.0074275558193524676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,512,0.011056000159846412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,5120,0.03869688842031691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,256,0.009690666364298927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,256,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,64,0.003552888830502828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,128,0.009692444569534725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,32,0.003583111282851961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,128,0.007339555356237624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,32,0.00889511075284746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,512,0.01108266661564509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,65536,0.3216408888498942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,4096,0.07353599866231282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,128,0.004042666819360522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,1024,0.021570665968788996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,51200,0.25076089964972603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,65536,0.17286845048268637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,16384,0.09150311019685532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,65536,0.1530053350660536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,1024,0.013431111143694984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,16384,0.041963554090923734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,64,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,10240,0.05307022068235609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,16384,0.047184000412623085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,12288,0.04503733250829908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,8192,0.04519199993875292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,10240,0.028767112228605483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,10240,0.043678220775392324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,7168,0.03804622093836466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,8192,0.03422133458985223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,6144,0.03399644295374552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,7168,0.02218311031659444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,7168,0.03070577647950914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,12288,0.07056888606813219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,51200,0.13082133399115667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,5120,0.028897778855429754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,4096,0.024645333488782246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,5120,0.017872000734011333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,5120,0.023617777559492324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,3584,0.021425777011447485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,4096,0.021236444513003033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,3584,0.014489778214030795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,3584,0.018223111828168232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,3072,0.018580444984965853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,3072,0.01316177762216992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,51200,0.12901333967844644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,3072,0.016908443636364408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,12288,0.039341333839628435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,2048,0.01331733332739936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,4096,0.01593866613176134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,2048,0.011069333387745751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,2048,0.013434666726324292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,1536,0.010875555376211802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,1536,0.009408888717492422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,1024,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,1536,0.011408000356621213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,2560,0.015861334072219003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,768,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,1024,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,6144,0.02971733278698391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,1024,0.009010666774378883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,768,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,2560,0.0140657772620519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,8192,0.02464177707831065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,768,0.008012444608741337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,256,0.003818666769398583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,512,0.007608888877762689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,512,0.00666844430896971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,256,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,256,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,128,0.00700800038046307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,128,0.005583999885453119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,32,0.0029235554652081597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,64,0.006091555373536215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,2560,0.014117333624098035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,32,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,65536,0.2523146735297309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,128,0.0031902222997612427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,65536,0.1476800044377645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,51200,0.19876710573832193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,65536,0.13248356183369955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,64,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,16384,0.06620799832873873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,51200,0.1124746667014228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,12288,0.05207822057935926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,16384,0.03752800159984165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,10240,0.04456266760826111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,512,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,12288,0.03359555535846286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,12288,0.04222666554980808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,8192,0.037463112009896174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,10240,0.02607022225856781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,10240,0.0374017788304223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,7168,0.03295644455485874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,8192,0.03102311160829332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,6144,0.02882400155067444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,7168,0.020267556111017864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,7168,0.026495110657480028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,6144,0.020005333754751418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,51200,0.1029706663555569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,16384,0.04132711225085788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,4096,0.02038755516211192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,5120,0.016295111841625638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,3584,0.017551110850440133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,5120,0.02034222251839108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,4096,0.019276445110638935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,3072,0.015636444091796875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,8192,0.02204088866710663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,3584,0.013445333474212222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,3584,0.01681066718366411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,2560,0.01316177762216992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,3072,0.014292443792025248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,3072,0.016407999727461074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,2560,0.011408000356621213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,5120,0.025382222400771245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,1536,0.009345778160625035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,2048,0.010028444230556488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,6144,0.024713777833514746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,1024,0.007875555919276344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,1536,0.008736888567606607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,4096,0.014690667390823364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,1536,0.01056977775361803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,1024,0.007715555528799693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,1024,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,512,0.004242666479614046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,768,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,768,0.007303111255168915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,256,0.003543111185232798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,512,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,512,0.006322666588756774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,128,0.0031795555518733132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,256,0.006613333192136552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,2048,0.012368888490729861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,64,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,128,0.00665155549844106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,32,0.0030462222380770575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,64,0.00592533333433999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,32,0.005967999911970562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,65536,0.21664178371429443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,2560,0.012528888881206512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,256,0.005676444619894028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,51200,0.17246577474806044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,128,0.005617777920431561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,16384,0.0594008896085951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,51200,0.10717599921756321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,51200,0.09837066464953953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,12288,0.04622222317589653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,16384,0.035747554567125105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,16384,0.0396364430586497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,10240,0.038231998682022095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,12288,0.028370665179358587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,12288,0.036269333627488874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,8192,0.03258400162061056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,65536,0.1265475617514716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,10240,0.024615110622511968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,10240,0.032441778315438166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,768,0.006717333363162146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,8192,0.026540444956885442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,6144,0.026035555534892615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,7168,0.023573334018389385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,2048,0.011378666592968835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,6144,0.01756177842617035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,6144,0.02068711154990726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,65536,0.14002133740319148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,5120,0.017887999614079792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,4096,0.01865688959757487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,5120,0.015810666812790763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,8192,0.021528000632921856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,3584,0.01611022154490153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,4096,0.013904889424641928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,7168,0.01964977714750502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,4096,0.016311999824311998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,3072,0.014392889208263822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,3584,0.012914666699038612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,3584,0.014173333843549093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,3072,0.01182755579551061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,6144,0.018226666582955253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,3072,0.013346666263209449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,2048,0.010657777388890585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,2560,0.010611555642551847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,1536,0.008963555925422246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,2560,0.011731555892361535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,2048,0.008984000318580205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,7168,0.029294222593307495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,1536,0.00850666645500395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,1536,0.00942311104800966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,768,0.005620444400442972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,1024,0.007536888950400882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,2560,0.012238221863905588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,5120,0.02269511090384589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,512,0.004842666702138053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,512,0.007027555671003129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,256,0.003979555434650845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,1024,0.0073520003093613525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,256,0.006311111152172089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,1024,0.0076657773719893555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,256,0.005650666852792104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,64,0.0031884445084465873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,128,0.00666933341158761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,768,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,768,0.006701333241330252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,64,0.00544888898730278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,65536,0.17548622025383842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,32,0.005621333503060871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,2048,0.011061333119869232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,51200,0.130949338277181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,65536,0.08924000130759345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,65536,0.07195466756820679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,128,0.0032071111102898917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,51200,0.06862933105892606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,12288,0.032807111740112305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,16384,0.027215111586782668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,12288,0.019239111079110038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,12288,0.02732444471783108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,10240,0.02398666739463806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,51200,0.056735111607445605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,16384,0.022943110929595098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,32,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,8192,0.01482311056719886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,8192,0.02012266715367635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,512,0.006267555471923616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,7168,0.01367555558681488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,6144,0.019107555349667866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,16384,0.04566399918662178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,10240,0.016564443707466125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,8192,0.02520977788501316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,6144,0.016573333077960543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,128,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,4096,0.014483556151390076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,5120,0.011778666741318174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,5120,0.014482667048772177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,7168,0.021714667479197185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,3584,0.012422222230169507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,4096,0.013434666726324292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,3072,0.011363555159833698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,3584,0.009869332942697737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,7168,0.01794755624400245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,3584,0.012112889024946423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,3072,0.00906133320596483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,3072,0.011394666300879585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,5120,0.016639111770523917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,2560,0.009926222264766693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,2048,0.007698666718271043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,2048,0.009334222310119206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,1536,0.007641777396202087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,4096,0.010750222537252637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,1024,0.005667555663320754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,1536,0.007358222372002072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,1536,0.007711999946170383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,2560,0.010094222095277574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,1024,0.006677333265542984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,1024,0.006747555401590135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,512,0.003910222401221593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,768,0.006676444576846228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,768,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,256,0.0032551110618644287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,512,0.006335999816656113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,512,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,128,0.0029013332807355454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,256,0.006021333237489064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,256,0.005477333234416113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,64,0.0028782222006056043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,128,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,768,0.004218666503826777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,32,0.0028773333049482773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,64,0.0053048887186580235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,2048,0.008660444782839881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,65536,0.14902755949232313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,32,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,51200,0.11623377270168728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,65536,0.08228088749779595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,65536,0.06606755654017131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,16384,0.041837334632873535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,51200,0.05405244562360975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,51200,0.0626479983329773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,12288,0.03289511137538486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,16384,0.021693333983421326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,6144,0.012768000364303589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,2560,0.008348444269763099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,12288,0.017895999881956313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,12288,0.026143999563323125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,8192,0.022135110365019903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,10240,0.0277173337009218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,8192,0.014255111416180929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,128,0.006315555423498154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,8192,0.019928889142142404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,7168,0.020620443754725985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,16384,0.025750221477614507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,10240,0.028088887532552082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,7168,0.017119111286269296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,5120,0.015836444165971544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,6144,0.012306667036480375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,6144,0.016238222519556682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,10240,0.022678222921159532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,5120,0.011054221954610614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,3584,0.0119991112086508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,4096,0.009943111075295342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,4096,0.012538666526476542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,3072,0.010724444356229572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,3584,0.009030222064918941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,3584,0.011566222541862063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,2560,0.009517333573765224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,3072,0.008010666403505538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,3072,0.010715555813577441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,2048,0.00833155545923445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,2560,0.007663110891977946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,2560,0.00906399968597624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,10240,0.015831111205948722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,1536,0.007299555672539606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,2048,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,1024,0.004907555464241239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,1536,0.007019555403126611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,1536,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,768,0.004231111043029361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,1024,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,1024,0.006312888943486744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,512,0.00388711111413108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,6144,0.01771111124091678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,512,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,512,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,4096,0.013722666435771518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,5120,0.013850666582584381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,256,0.0059004442559348206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,128,0.00309599998096625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,128,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,64,0.002941333378354708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,768,0.006326222171386083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,32,0.0028906667398081887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,64,0.005323555734422471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,32,0.0052977779673205475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,65536,0.12323110633426243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,2048,0.008381333616044786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,256,0.0052986666560173035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,51200,0.09682399696773952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,65536,0.0756817791197035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,65536,0.06377510892020331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,128,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,51200,0.05151289039187961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,7168,0.013153778182135688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,16384,0.02113688819938236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,16384,0.02367377777894338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,256,0.003195555466744635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,12288,0.02369333306948344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,10240,0.015565334094895257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,10240,0.02120888895458645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,8192,0.021546666820844013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,16384,0.03600000010596381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,8192,0.01386044422785441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,7168,0.018975110517607797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,51200,0.05654578076468574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,768,0.00630044440428416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,6144,0.01699022286468082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,7168,0.012864888542228274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,5120,0.015325332681337992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,6144,0.012015110916561551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,10240,0.024691555235120986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,6144,0.014811555544535318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,4096,0.011807999677128263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,5120,0.010750222537252637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,3584,0.010212444596820408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,4096,0.009151111046473185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,4096,0.012088889049159156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,3072,0.010371555884679159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,3584,0.008373333348168267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,3584,0.01106844428512785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,8192,0.020059555768966675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,12288,0.028559111886554297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,2560,0.0090506664580769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,3072,0.010273777776294285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,2560,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,2560,0.00868622213602066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,5120,0.013146667016877068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,1536,0.006888888776302338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,1536,0.007292444507280986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,3072,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,1536,0.006724444528420766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,7168,0.016191111670600045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,12288,0.017278222574128043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,768,0.0042444442709287005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,1024,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,1024,0.0063324446479479475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,2048,0.007330666813585494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,768,0.006321777900060018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,2048,0.007831999825106727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,256,0.0032008888406885993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,512,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,512,0.005595555735958948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,128,0.002883555574549569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,256,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,1024,0.004741333425045013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,128,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,2048,0.008007110820876228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,128,0.004997333304749595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,32,0.0025457777082920074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,512,0.0038648889296584656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,32,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,768,0.0059075554211934405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,51200,0.08521511157353719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,65536,0.0614924430847168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,16384,0.031927112076017596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,256,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,51200,0.04948177933692932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,12288,0.0259226659933726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,16384,0.02032177812523312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,64,0.0028657778683635923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,16384,0.02126755482620663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,10240,0.022388445006476507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,12288,0.016897777716318767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,8192,0.019535111056433786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,10240,0.015041778484980265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,64,0.005296000176005893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,10240,0.019369777705934312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,8192,0.013453333742088742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,65536,0.06716710991329618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,6144,0.01574488812022739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,7168,0.012452444268597497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,7168,0.014827556080288358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,5120,0.014076444837782117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,6144,0.011457777685589261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,6144,0.014126222994592456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,4096,0.011735111474990845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,5120,0.00983199973901113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,5120,0.012167111039161682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,3584,0.010507555471526252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,4096,0.00868711123863856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,7168,0.017437333861986797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,3072,0.009391999906963771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,8192,0.01720888912677765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,3072,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,3072,0.00996088898844189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,2560,0.0084906667470932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,2560,0.007687999970383114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,2048,0.0075902218619982404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,12288,0.021351110604074266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,65536,0.10652799738778009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,1536,0.006180444525347815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,2048,0.007109333243634965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,3584,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,1024,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,1536,0.006640000061856375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,1536,0.0070017774899800616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,51200,0.05111822154786852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,1024,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,768,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,2560,0.00793333351612091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,4096,0.01145866678820716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,2048,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,256,0.0031911111954185697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,512,0.006116444451941385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,256,0.005965333431959152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,1024,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,256,0.005285333428117964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,128,0.005980444451173146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,768,0.005971555494599872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,512,0.005430222385459476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,32,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,768,0.004031110968854693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,32,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,65536,0.08613777822918361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,512,0.003559111307064692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,65536,0.06082666582531399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,65536,0.059311111768086754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,64,0.0052791109515561005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,3584,0.010500444306267632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,51200,0.04786488744947645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,51200,0.04667377803060743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,128,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,12288,0.021864000293943617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,16384,0.019119999474949308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,64,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,10240,0.018836443622907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,10240,0.017682666579882305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,16384,0.02758933272626665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,8192,0.01722666621208191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,10240,0.014595554934607612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,16384,0.019641778535313077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,7168,0.013767110804716745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,8192,0.012757333616415659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,8192,0.01605244477589925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,6144,0.012467555701732635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,7168,0.011778666741318174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,12288,0.02055022286044227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,5120,0.011048888994587792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,6144,0.01054044481780794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,6144,0.01331822243001726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,4096,0.010107555323176915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,5120,0.011515555282433828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,4096,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,3584,0.008736888567606607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,128,0.004934222333961063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,51200,0.06730666425493029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,3584,0.009376889301670922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,3584,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,2560,0.007340444458855524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,3072,0.007734222544564142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,5120,0.009534222384293875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,2048,0.006160000132189856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,2560,0.007361777954631382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,2560,0.0073831114504072405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,1536,0.004573333180612988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,4096,0.01109777804878023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,2048,0.007138667007287343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,1024,0.004231111043029361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,1536,0.006671111202902264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,12288,0.016211556063758004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,1536,0.006684444430801604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,7168,0.015817777978049386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,1024,0.0063164445261160536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,512,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,768,0.006267555471923616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,768,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,3072,0.008281778130266402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,256,0.003185777821474605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,512,0.0052915554907586836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,128,0.0028515555378463534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,256,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,2048,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,256,0.004946666873163647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,128,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,32,0.0028435554769304064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,768,0.003859555555714501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,64,0.004951999833186467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,65536,0.07773955663045247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,1024,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,32,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,51200,0.060524443785349526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,512,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,65536,0.0579342246055603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,16384,0.023731556203630235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,51200,0.04682222339842054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,12288,0.018915555543369718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,64,0.0028488888508743713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,128,0.004982222285535601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,12288,0.01588711142539978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,10240,0.017525333497259352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,51200,0.04474400149451362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,10240,0.014399111270904541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,16384,0.01924977699915568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,16384,0.018351111147138808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,7168,0.013792888985739814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,8192,0.012440889245933957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,7168,0.011400889191362592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,7168,0.013447110851605734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,65536,0.05820444557401869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,6144,0.01293511109219657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,8192,0.014993778533405729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,5120,0.01070666644308302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,5120,0.009353777600659264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,4096,0.009522666533788046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,5120,0.010741333166758219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,4096,0.008361777497662438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,6144,0.011731555892361535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,4096,0.009858667022652095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,3584,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,3072,0.007635555333561367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,6144,0.010391111175219217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,3072,0.007666666474607255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,2560,0.006766222417354584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,10240,0.016895999511082966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,2560,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,2048,0.005576000031497743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,3584,0.008259555531872643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,2048,0.007024889190991719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,2048,0.0069546666410234236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,1536,0.004561777744028303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,3072,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,12288,0.019288000133302476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,1024,0.003944888710975647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,1536,0.006702222343948152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,1536,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,2560,0.007542221910423703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,1024,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,1024,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,512,0.003506666670242945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,768,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,768,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,512,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,512,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,3072,0.007976888782448238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,256,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,128,0.002834666727317704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,768,0.003559999995761447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,64,0.00254488881263468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,64,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,8192,0.017324444320466783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,32,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,65536,0.06287022431691487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,128,0.005623111294375525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,65536,0.05718666977352566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,65536,0.05490844448407491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,51200,0.04598844382497999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,256,0.0028880000528362063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,16384,0.020439111524158053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,16384,0.01923466722170512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,12288,0.017321777012613084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,3584,0.00852177788813909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,128,0.004941333499219683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,12288,0.017818666166729398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,12288,0.015863110621770222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,10240,0.013796444568369122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,32,0.0028284444577164124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,8192,0.015173332558737861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,8192,0.011686222420798408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,7168,0.012413333687517377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,51200,0.05072089036305746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,7168,0.012502222425407834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,10240,0.016861332787407767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,6144,0.011223999990357293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,7168,0.010716444088353051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,5120,0.010037333601050908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,6144,0.009701333112186855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,51200,0.04319555560747782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,6144,0.012120889292822944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,4096,0.009213333328564962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,5120,0.009054222040706212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,3584,0.008083555433485243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,5120,0.010033778018421596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,4096,0.008416888614495596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,16384,0.01794933279355367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,3072,0.007332444190979004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,3584,0.008006222546100616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,2560,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,3072,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,2560,0.0069324444565508105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,2048,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,10240,0.015950222810109455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,2048,0.006989333364698622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,1536,0.004517333375083076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,4096,0.008647999829716152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,3072,0.007672888537247975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,1024,0.0038684445122877755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,2560,0.007351111206743453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,1024,0.005766222046481238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,768,0.003526222374704149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,1024,0.006303999986913469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,256,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,768,0.005988444305128521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,768,0.005620444400442972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,3584,0.007985778152942657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,512,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,1536,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,256,0.0058062221441004015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,128,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,256,0.005193777796294955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,128,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,64,0.0028977776981062363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,1536,0.00666222224632899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,32,0.0025173332542181015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,32,0.004937777916590373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,65536,0.05743555227915446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,128,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,65536,0.05271466573079427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,65536,0.05634222428003947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,512,0.005921777751710679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,256,0.003160888950030009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,16384,0.019161777363883126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,51200,0.04524266719818115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,51200,0.041454222467210554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,12288,0.01554844445652432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,16384,0.018953777021831937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,16384,0.017196445001496207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,10240,0.014476444986131458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,12288,0.01554666625128852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,512,0.0032115555885765287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,10240,0.015166223049163818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,64,0.0052879999081293745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,8192,0.013140444126394061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,8192,0.011736888852384357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,7168,0.01180622229973475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,2048,0.006688000013430913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,10240,0.013306666579511432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,51200,0.04515555500984192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,6144,0.011061333119869232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,6144,0.009703999592198266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,5120,0.01038577738735411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,5120,0.00850844466023975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,8192,0.013707555830478668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,4096,0.008906666603353288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,5120,0.009359111388524374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,7168,0.010370666782061258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,8192,0.012730666332774691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,4096,0.008630221916569604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,4096,0.007970666719807519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,3072,0.006692444284756978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,6144,0.010781333678298526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,2560,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,3072,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,2560,0.007625777688291337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,12288,0.01550044450494978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,7168,0.011902222202883827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,2048,0.0063457778758472866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,1536,0.004600000050332812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,2048,0.007035555938879649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,3584,0.007636444436179266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,1024,0.0038968887594011095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,1536,0.0058133333093590206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,1024,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,1024,0.005584888988071018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,768,0.0035377778112888336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,768,0.006274666637182236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,512,0.003192000091075897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,768,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,3584,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,256,0.00318755561278926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,512,0.0052675555149714155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,256,0.005973333285914526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,256,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,2560,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,2048,0.005066666752099991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,64,0.0025066667132907445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,128,0.005616888817813661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,128,0.004940444396601783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,64,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,65536,0.04985333482424418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,32,0.004931555440028508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,1536,0.00664444433318244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,51200,0.037019554111692644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,65536,0.05545599924193489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,65536,0.05008622341685825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,16384,0.016520889268981088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,51200,0.044848889112472534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,51200,0.04024089044994778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,32,0.0028773333049482773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,16384,0.016688000824716356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,12288,0.014127110441525778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,12288,0.015839111473825242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,10240,0.013320888910028668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,3072,0.007882666256692674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,8192,0.012510221865442065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,10240,0.012432888978057437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,128,0.0028728888266616394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,10240,0.014418666561444601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,7168,0.010995555255148145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,8192,0.011065777805116443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,3584,0.008031999899281396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,7168,0.01038044442733129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,7168,0.010420444111029306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,5120,0.009743111001120673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,6144,0.009479111267460717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,512,0.005991111199061076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,6144,0.009370666411187913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,5120,0.009020444419648912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,5120,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,3584,0.007294221884674496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,4096,0.008332444561852349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,3584,0.00796888851457172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,8192,0.011730666789743634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,12288,0.014478221535682678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,3072,0.007673777639865875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,3072,0.007037333316273159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,4096,0.008248888783984715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,2560,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,2048,0.005239999956554837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,4096,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,2048,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,3584,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,2048,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,1536,0.006635555376609166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,1024,0.003941333128346337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,1024,0.00562755556570159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,1024,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,768,0.003584889074166616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,3072,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,16384,0.018576888574494254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,768,0.005997333261701796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,768,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,512,0.005244444641802046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,512,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,256,0.0029031110720502008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,1536,0.004294222427739037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,6144,0.01035911093155543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,128,0.0028479999552170434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,256,0.005622222191757626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,256,0.005092444519201915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,512,0.0035413333939181436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,64,0.0025155554629034465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,32,0.002519999941190084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,64,0.004965333475006951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,65536,0.03624533282385932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,65536,0.04861866765552097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,2560,0.005639111002286275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,128,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,51200,0.02914311157332526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,2560,0.006613333192136552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,51200,0.038740442858801946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,16384,0.01720977822939555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,12288,0.012774222426944308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,1536,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,12288,0.014183110660976835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,128,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,10240,0.011447110937701331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,12288,0.014473777678277759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,8192,0.010388444695207808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,10240,0.012123555772834353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,8192,0.01037688884470198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,8192,0.010794666906197866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,16384,0.01627200014061398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,7168,0.010421333213647207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,65536,0.055105778906080455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,7168,0.009371555513805812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,6144,0.009730666875839233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,5120,0.008624888956546783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,51200,0.04442755712403191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,5120,0.008904000123341879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,7168,0.009679999616410997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,6144,0.009856888817416297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,4096,0.0074515557951397365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,3584,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,6144,0.009018667042255402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,16384,0.01421333352724711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,3584,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,3584,0.007340444458855524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,5120,0.008169777691364288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,4096,0.008012444608741337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,3072,0.007751999629868402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,3072,0.0069822221994400024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,2048,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,2560,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,2560,0.006670222100284364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,1536,0.004552888787455029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,2048,0.0070017774899800616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,2048,0.0063182223174307085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,1024,0.00387999994887246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,1536,0.006656888872385025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,32,0.00498488876554701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,1536,0.00600888869828648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,1024,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,1024,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,2560,0.00590133335855272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,512,0.0032115555885765287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,768,0.005983110931184556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,768,0.005362666729423735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,512,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,512,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,256,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,4096,0.008387555678685507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,128,0.005679111099905438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,10240,0.012347555822796293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,64,0.0028453332682450614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,768,0.003543111185232798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,64,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,256,0.0030257778449190986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,128,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,256,0.005004444470008214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,51200,0.02160888910293579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,65536,0.054423110352622144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,128,0.004939555707905027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,16384,0.011785777906576792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,51200,0.043839110268486865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,51200,0.0377422206931644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,16384,0.016134222348531086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,16384,0.01583377851380242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,12288,0.013455111119482251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,12288,0.01277955538696713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,65536,0.04670400089687771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,10240,0.012130666938092975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,10240,0.011067555182509951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,3072,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,8192,0.010655110908879174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,12288,0.011360888679822287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,8192,0.00979466653532452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,65536,0.025947555899620056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,6144,0.008672888908121321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,7168,0.010064000056849586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,7168,0.009407111340098912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,6144,0.009402666654851701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,5120,0.008033778104517195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,8192,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,32,0.0028799999919202593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,4096,0.00738844441043006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,5120,0.009015111459626092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,7168,0.009712888962692684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,3584,0.00692622239391009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,4096,0.007348444726732042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,3584,0.008024888734022776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,3072,0.0063288890653186375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,6144,0.008667555948098501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,3072,0.006996444529957241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,2560,0.005626666463083691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,5120,0.008173333273993598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,32,0.004931555440028508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,2048,0.006400888992680444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,2560,0.0073520003093613525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,10240,0.01036533299419615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,2048,0.007050666544172499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,2048,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,1536,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,3072,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,1024,0.006029333505365584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,768,0.004257777912749184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,2560,0.00656444455186526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,768,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,3584,0.007151111132568783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,768,0.0052764444715446895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,512,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,1024,0.004559111264016894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,512,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,256,0.005611555443869696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,128,0.0029288888391521242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,1024,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,128,0.005783111270931032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,64,0.0025271111064487034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,1536,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,512,0.003552888830502828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,256,0.0032293332947625052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,1536,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,768,65536,0.023764444722069636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,65536,0.054096887509028115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,4096,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,768,51200,0.019242667489581637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,51200,0.0432515541712443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,768,16384,0.011399110986126794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,64,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,16384,0.014550222290886773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,768,12288,0.010791111323568555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,16384,0.015843555331230164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,256,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,768,10240,0.009690666364298927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,12288,0.012122666670216454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,128,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,10240,0.011815110842386881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,51200,0.03721333212322659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,8192,0.010770666930410596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,768,7168,0.009392889009581672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,32,0.004947555561860402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,7168,0.010296888649463654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,768,6144,0.008593777815500895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,12288,0.013447999954223633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,6144,0.009396444592210982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,768,5120,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,768,8192,0.008623110751310984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,5120,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,8192,0.009746666583749983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,768,4096,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,4096,0.007978666987684038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,768,3584,0.006787555499209299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,65536,0.04634933339224922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,3584,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,768,3072,0.0063164445261160536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,10240,0.011063999599880643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,3072,0.007632888853549957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,768,2560,0.005576000031497743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,3072,0.00702311098575592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,2560,0.007318221860461765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,768,2048,0.0052426668504873914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,6144,0.008924444516499838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,32,0.0028782222006056043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,768,1536,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,2048,0.007019555403126611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,5120,0.008016889293988546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,1536,0.006650666809744305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,768,1024,0.004222222086456087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,1536,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,768,768,0.0035457776652442086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,1024,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,2560,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,1024,0.00563466673096021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,768,0.005984888722499211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,768,0.005396444350481033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,768,256,0.0032195556494924757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,512,0.0053119998839166425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,256,0.005759999983840519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,256,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,7168,0.009368889033794403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,768,512,0.0032008888406885993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,128,0.005712888720962737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,512,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,64,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,768,32,0.0028568889117903183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,4096,0.007473777565691207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,32,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,768,64,0.00253866674999396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,65536,0.045533332559797496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,512,51200,0.016192888220151264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,768,128,0.0028328889360030494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,51200,0.03860800133811103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,512,16384,0.010468444062603844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,512,65536,0.01731911136044396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,16384,0.015862221519152325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,2048,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,512,12288,0.010067555639478896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,65536,0.0513937771320343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,512,10240,0.00869422240389718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,12288,0.0123511114054256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,10240,0.012103999654452005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,51200,0.036679999695883855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,8192,0.01089511149459415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,8192,0.009791110952695211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,16384,0.014521777629852295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,3584,0.007286222444640265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,512,6144,0.008354666332403818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,7168,0.009374221993817223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,7168,0.010132444401582083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,512,5120,0.007671111159854465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,6144,0.009359111388524374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,6144,0.00870400004916721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,5120,0.008956444760163626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,10240,0.011050666371981302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,5120,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,512,3584,0.006924444602595435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,4096,0.008095111283991072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,512,7168,0.009016888837019602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,3584,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,3584,0.007361777954631382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,3072,0.007638221813572778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,12288,0.013110222087966071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,512,8192,0.008044444852405125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,512,4096,0.007328888608349695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,512,2048,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,2560,0.007351111206743453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,2048,0.007010666860474481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,512,1536,0.004571555389298333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,128,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,1536,0.006666666517655055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,512,2560,0.006311999840868845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,2560,0.0063031112982167145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,1024,0.006181333214044571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,4096,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,2048,0.006019555446174409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,768,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,512,512,0.0032186667538351486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,3072,0.006685333533419504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,512,0.005913777897755305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,512,256,0.002887111157178879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,1024,0.005636444522274866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,256,0.00562755556570159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,256,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,512,3072,0.006505777852402792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,128,0.005759999983840519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,768,0.0052728888889153796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,1536,0.005963555640644497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,512,32,0.002554666664865282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,512,0.004938666605287128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,256,65536,0.017147555947303772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,32,0.004965333475006951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,512,768,0.0035502223504914176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,65536,0.046712001164754234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,256,51200,0.01480888823668162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,51200,0.03738488753636678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,512,64,0.0026373333401150177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,256,16384,0.009371555513805812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,512,1024,0.004002666721741359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,256,12288,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,16384,0.015857777661747403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,64,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,256,10240,0.007624000310897827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,12288,0.01312800000111262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,12288,0.012085333466529846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,256,8192,0.0069520001610120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,10240,0.011916444533401065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,512,128,0.002550222186578645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,10240,0.011063999599880643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,8192,0.0107351111041175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,8192,0.010037333601050908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,128,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,7168,0.009379555781682333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,6144,0.009363555245929295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,16384,0.014472888575659858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,5120,0.008016000191370646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,256,7168,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,256,4096,0.006321777900060018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,5120,0.008841777841250101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,256,6144,0.007227555745177799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,4096,0.00831911133395301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,256,3584,0.006112888869312074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,6144,0.00868622213602066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,3584,0.0069919998447100324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,3584,0.007970666719807519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,256,3072,0.005990222096443176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,65536,0.04232088724772135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,256,2560,0.005893333504597346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,3072,0.006713777780532837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,7168,0.010043555663691627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,2560,0.007328888608349695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,4096,0.007332444190979004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,256,5120,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,256,1536,0.004590222405062782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,1536,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,1536,0.006635555376609166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,256,1024,0.003608889049953885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,51200,0.03485066692034403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,2560,0.006260444306664997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,256,768,0.003559111307064692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,1024,0.005567111074924469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,2048,0.006299555715587404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,256,512,0.0032026666320032547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,768,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,768,0.0052915554907586836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,256,256,0.0028488888508743713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,512,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,3072,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,256,128,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,512,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,256,64,0.0028693332440323303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,256,0.004946666873163647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,256,32,0.0028693332440323303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,256,2048,0.005622222191757626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,128,65536,0.014499555031458536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,64,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,2048,0.00703288863102595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,128,51200,0.012461333639091916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,65536,0.04178399840990702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,51200,0.0373599992858039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,256,0.005655111124118169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,128,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,128,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,128,12288,0.007640000018808577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,1024,0.006309333360857434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,32,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,128,10240,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,12288,0.012051555845472546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,10240,0.012090666426552666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,10240,0.011071110765139261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,128,16384,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,8192,0.010728889041476779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,8192,0.009880888793203566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,16384,0.01585422290696038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,7168,0.010107555323176915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,65536,0.046219554212358266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,128,8192,0.006584888945023219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,6144,0.009364444348547194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,128,5120,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,128,7168,0.006302222195598815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,16384,0.014196443888876172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,128,4096,0.005558222118351195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,5120,0.00870400004916721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,5120,0.007667555577225155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,128,6144,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,128,3584,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,4096,0.007181333170996771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,128,3072,0.005636444522274866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,6144,0.008839999636014303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,3584,0.006949333681000604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,128,2560,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,2560,0.006637333167923822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,12288,0.01348622226052814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,4096,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,128,2048,0.005621333503060871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,3584,0.00794222205877304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,2048,0.0069573331210348345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,51200,0.03401155604256524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,2048,0.0060159998635451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,128,1024,0.003607999947335985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,1536,0.006660444455014334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,1536,0.005986666513813867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,1024,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,1024,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,768,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,7168,0.009011555876996782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,128,512,0.0031999999450312722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,2560,0.007300444775157505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,128,256,0.0029226665695508323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,512,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,128,1536,0.004575110971927643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,512,0.0052826665341854095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,256,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,256,0.005037333402368757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,3072,0.007673777639865875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,128,64,0.0026391111314296722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,128,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,128,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,64,65536,0.011186666786670685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,64,0.004880889008442561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,32,0.004600888739029567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,64,51200,0.01073066641887029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,768,0.005542222410440445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,65536,0.042134222057130605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,128,128,0.002512000087234709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,64,12288,0.007290666302045186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,3072,0.006640888750553131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,64,10240,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,128,32,0.002537777854336633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,128,768,0.003625777860482534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,12288,0.012097777591811286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,64,7168,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,10240,0.01074488874938753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,64,6144,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,8192,0.009725333087974125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,64,16384,0.007000889215204451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,7168,0.009199110998047723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,64,4096,0.005582222094138463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,51200,0.0336133340994517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,64,3584,0.005619555711746216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,4096,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,64,3072,0.005895111295912001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,3584,0.006981333096822103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,64,2560,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,3072,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,64,2048,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,2560,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,64,1536,0.004568888909286923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,6144,0.008344000412358178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,64,1024,0.0038817777401871155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,5120,0.008032889001899296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,64,768,0.0035342222286595237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,64,512,0.0031911111954185697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,1536,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,64,256,0.0028657778683635923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,1024,0.005283555636803309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,768,0.00562755556570159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,64,128,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,64,64,0.002519999941190084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,256,0.004597333156400257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,64,32,0.0025395556456512874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,64,5120,0.006015111174848344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,32,65536,0.011101333631409539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,64,0.004598222259018156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,32,51200,0.010386666489972008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,32,0.004598222259018156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,32,16384,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,65536,0.04220622115665012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,32,12288,0.00702044450574451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,16384,0.014157333307796054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,512,0.00526577772365676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,12288,0.012085333466529846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,32,8192,0.0063235556913746735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,128,0.004939555707905027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,32,7168,0.006640888750553131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,16384,0.014478221535682678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,32,6144,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,7168,0.009069333473841349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,32,5120,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,64,8192,0.006792888873153263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,32,4096,0.005828444328573015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,6144,0.00866844422287411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,32,10240,0.006299555715587404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,5120,0.008016889293988546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,32,3072,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,4096,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,32,2560,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,10240,0.010732444624106089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,3584,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,32,2048,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,32,1536,0.004326222257481681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,2048,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,32,1024,0.003551111039188173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,2048,0.005964444329341252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,51200,0.03360266817940606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,1536,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,32,512,0.003195555466744635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,1024,0.005586666779385672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,32,256,0.0028568889117903183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,768,0.005249777601824866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,32,128,0.0026853332916895547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,512,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,256,0.005038222091065513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,3072,0.006945778098371293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,8192,0.009741333623727163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,32,32,0.0025457777082920074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,64,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,32,768,0.0032808888289663526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,32,0.005039999882380168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,32,3584,0.006165333506133821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,16384,0.32385865847269696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,51200,0.4882604281107585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,51200,0.9724915822347006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,16384,0.16325155893961588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,2560,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,12288,0.24749689631991914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,16384,0.13400444719526503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,10240,0.18816710842980278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,51200,0.41466487778557676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,12288,0.10309955808851455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,10240,0.1508293284310235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,10240,0.086081776354048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,10,32,64,0.00254488881263468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,128,0.004834666848182678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,8192,0.07014133532842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,6144,0.11507911152309841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,7168,0.11262310875786676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,6144,0.055184000068240695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,12288,0.1438551081551446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,8192,0.15488444434271917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,5120,0.09669599930445354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,4096,0.0795217752456665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,7168,0.132968889342414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,5120,0.07484177748362224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,5120,0.047592888275782265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,3584,0.06885066959593031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,4096,0.062126219272613525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,7168,0.06291021903355916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,3072,0.05993600024117363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,3584,0.042394667863845825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,2560,0.05040533343950907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,6144,0.0827653341823154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,3072,0.04753777715894911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,2048,0.04190399911668566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,2560,0.02715911136733161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,2560,0.04220799936188591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,1536,0.03223289052645365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,8192,0.11402132776048447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,2048,0.03358488943841722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,4096,0.039015998442967735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,1024,0.023377777801619634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,1536,0.01925155520439148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,768,0.018406222263971966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,1536,0.027098667290475633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,1024,0.015258666541841296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,512,0.013487999637921652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,768,0.015174221661355762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,256,0.008455110920800103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,768,0.016131555040677387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,512,0.012433778080675336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,128,0.006370666540331311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,256,0.010409778190983666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,256,0.010034666293197209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,64,0.005790222022268508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,128,0.010755555497275459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,32,0.005967999911970562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,64,0.008725333544943068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,3584,0.05531466669506497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,32,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,1024,0.01963644391960568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,512,0.012915555801656512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,2048,0.023131555981106226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,65536,0.5146080123053657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,51200,0.39014400376213926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,65536,1.0903537538316515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,51200,0.8523715337117513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,16384,0.28226134512159556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,3072,0.03141599893569946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,12288,0.20970666408538818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,16384,0.1304239961836073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,65536,0.4346640110015869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,12288,0.11588622464074029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,12288,0.08272000153859456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,10240,0.17485600047641328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,8192,0.14190577136145696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,10240,0.07220621903737386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,51200,0.36615199512905544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,7168,0.12466577688852946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,8192,0.1059795551829868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,6144,0.10733066664801703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,7168,0.07640888955858019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,7168,0.05065244436264038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,5120,0.08921333154042561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,8192,0.057336886723836265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,6144,0.0769617756207784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,6144,0.04427466789881388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,4096,0.0731395549244351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,5120,0.03940355446603563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,16384,0.11114133728875054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,4096,0.03208266695340475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,128,0.008040889269775814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,10240,0.0988088912434048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,3584,0.029779556724760268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,3584,0.045216000742382474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,3072,0.03135822216669718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,3072,0.04042488998836941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,2048,0.03903199897872077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,2560,0.022974222898483276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,2560,0.03379822108480666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,1536,0.02991644541422526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,2048,0.019688000281651814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,2048,0.028224888775083754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,1024,0.02197066611713833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,1536,0.022216889593336318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,5120,0.060397333568996854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,768,0.016163556112183463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,1024,0.013462222284740873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,1024,0.016188444362746347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,512,0.011676444775528379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,768,0.01350933313369751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,768,0.014064888159434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,512,0.011057777537239922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,256,0.007585778004593319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,3584,0.06411733229955037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,4096,0.051721778180864125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,256,0.008685333033402761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,128,0.004247111164861255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,2560,0.04672444529003567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,64,0.003567111161020067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,128,0.009691555466916826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,32,0.0035253332720862497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,128,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,64,0.008004444340864817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,512,0.010891555911964841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,65536,0.3222951094309489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,51200,0.2504568894704183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,256,0.009766221874290043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,65536,0.15610667069753012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,16384,0.09085066450966729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,51200,0.11666488647460938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,32,0.008801778157552084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,16384,0.04768088791105482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,16384,0.04189866781234741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,3072,0.05583288934495714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,12288,0.03324800067477756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,10240,0.05285955468813578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,8192,0.044171555174721606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,10240,0.04072177741262648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,51200,0.13069866763220891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,7168,0.03813599877887302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,8192,0.024350222614076402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,8192,0.0358257790406545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,7168,0.022272000710169475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,1536,0.01683288812637329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,12288,0.052470223771201245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,6144,0.02000177734427982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,6144,0.027497778336207073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,10240,0.029100444581773546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,5120,0.0177208897140291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,4096,0.023964444796244305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,5120,0.024548444483015273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,3584,0.021029333273569744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,4096,0.01570044457912445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,4096,0.021617778473430212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,6144,0.033245331711239286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,3584,0.01959466603067186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,3072,0.016322667400042217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,3072,0.017065778374671936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,2560,0.015820443630218506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,2560,0.012517333030700684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,2560,0.01538222200340695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,12288,0.07089066505432129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,3072,0.018486221631368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,65536,0.17355199654897055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,2048,0.011037333144081963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,5120,0.02872177627351549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,1024,0.008257778154479133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,1536,0.011067555182509951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,1024,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,2048,0.012986666626400419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,768,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,768,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,512,0.004874666531880696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,7168,0.03154666556252374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,512,0.007635555333561367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,512,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,256,0.003552000141806073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,256,0.0069955554273393415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,128,0.0032248888164758682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,256,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,1024,0.008760000268618265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,64,0.0028408887899584244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,128,0.006992888947327931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,32,0.002891555635465516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,3584,0.014489778214030795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,1536,0.010724444356229572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,64,0.00600000015563435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,65536,0.2520044379764133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,65536,0.15009421772427028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,65536,0.1342408921983507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,51200,0.1121422184838189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,128,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,16384,0.06672000222735934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,1536,0.009528000321653154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,16384,0.04128177629576789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,12288,0.05189244283570183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,51200,0.19843022028605142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,10240,0.044701334502961904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,12288,0.04451555675930447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,10240,0.029694222741656836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,51200,0.10671644740634495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,8192,0.03730666637420654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,8192,0.02179466684659322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,16384,0.03758044375313653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,8192,0.033977776765823364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,7168,0.020287111401557922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,12288,0.02970222300953335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,6144,0.028888887829250757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,6144,0.018366222580273945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,5120,0.025294222765498694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,10240,0.033313777711656355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,32,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,7168,0.032778667079077825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,5120,0.021201777789327834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,5120,0.01629066632853614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,7168,0.026774222652117412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,3584,0.01761511133776771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,4096,0.01927911076280806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,2048,0.013616888887352414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,3072,0.015472888946533203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,3584,0.017588444881969027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,3072,0.012429333395428128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,4096,0.02030044462945726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,2560,0.01313777764638265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,2560,0.011142222417725457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,2048,0.011620444556077322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,2560,0.014118222726715935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,2048,0.00979288915793101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,3584,0.013533333109484779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,2048,0.01256622208489312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,1536,0.008665777742862701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,1024,0.007674666742483775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,3072,0.015615110596021017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,1024,0.007684444387753804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,4096,0.014480888843536377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,768,0.007669332954618666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,768,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,768,0.00795555528667238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,1536,0.009376000199053023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,256,0.0035377778112888336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,512,0.007334222396214803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,512,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,128,0.0031973332580592898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,256,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,768,0.006436444405052397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,128,0.006611555400821898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,128,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,32,0.003195555466744635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,512,0.004228444563017951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,32,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,65536,0.2169262170791626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,256,0.006856888946559694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,65536,0.14510311020745173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,51200,0.171888894504971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,65536,0.12841510772705078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,64,0.0032017777363459268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,1024,0.008204444415039485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,16384,0.05925510989295112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,12288,0.04630400074852837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,51200,0.10718933078977798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,6144,0.02644177774588267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,16384,0.039155556095971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,10240,0.03803999887572394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,12288,0.028215110301971436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,12288,0.03632888860172696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,8192,0.03209244542651706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,10240,0.028188443846172754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,1536,0.01039644413524204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,8192,0.021017778250906203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,8192,0.02605955633852217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,6144,0.025905778010686237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,7168,0.02272622287273407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,51200,0.10316889153586493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,6144,0.01704266667366028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,64,0.005617777920431561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,5120,0.01571644511487749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,4096,0.01862044466866387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,10240,0.03124622172779507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,4096,0.014012444350454541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,4096,0.016184889607959323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,3584,0.015850666496488783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,3584,0.012973333398501078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,6144,0.020090666082170274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,5120,0.022513777017593384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,3072,0.013448889056841532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,2560,0.012338666452301873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,3072,0.013341333303186627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,7168,0.029020445214377508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,2048,0.010776888993051318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,2560,0.011688888900809817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,2560,0.01276622215906779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,1536,0.009034666750166152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,2048,0.010049777726332346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,2048,0.01072799993885888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,1024,0.007641777396202087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,16384,0.03596355517705282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,1536,0.008269333177142674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,768,0.005759999983840519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,1024,0.00742222203148736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,1024,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,512,0.004896000027656555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,768,0.006763555523422029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,512,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,512,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,256,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,1536,0.00906311141120063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,128,0.0032257777121331957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,256,0.006692444284756978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,5120,0.017882666654056974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,64,0.0032071111102898917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,128,0.00534400012758043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,256,0.0038693332009845306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,32,0.0028808888875775864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,64,0.00563733321097162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,3072,0.014342221948835584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,32,0.005587555468082428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,768,0.0073902226156658595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,128,0.006665777828958299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,65536,0.08894044160842896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,51200,0.1304559972551134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,16384,0.045231110519833036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,7168,0.01901155544651879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,51200,0.05790399842792087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,12288,0.032602667808532715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,16384,0.02293599976433648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,16384,0.027591110931502447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,10240,0.02751911183198293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,12288,0.018372444642914664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,12288,0.027613333529896204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,8192,0.024971556332376268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,65536,0.0705439978175693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,10240,0.0236799998415841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,7168,0.02163822286658817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,8192,0.01477599971824222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,8192,0.0203813330994712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,51200,0.06879022386338976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,7168,0.013802666631009845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,7168,0.01829155617290073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,5120,0.016524444023768108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,3584,0.014873777826627096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,4096,0.014475555883513557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,5120,0.011738667057620155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,5120,0.01458755632241567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,10240,0.016391999191708036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,4096,0.010647111468844943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,4096,0.0129457778400845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,3072,0.011671110987663269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,3584,0.009926222264766693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,6144,0.019122666782803006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,3072,0.011061333119869232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,3072,0.009015111459626092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,65536,0.17381066746181914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,2048,0.008360889222886827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,2560,0.008360000120268928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,2560,0.010372444159454769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,1536,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,3584,0.012693333129088083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,2048,0.009101333717505137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,1024,0.005497777627574072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,1536,0.007329777710967594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,1536,0.007596444752481248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,768,0.004247111164861255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,1024,0.006670222100284364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,2560,0.009929777847396003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,512,0.0035413333939181436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,768,0.00664444433318244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,6144,0.012655110822783576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,768,0.006301333506902059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,256,0.003192000091075897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,512,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,128,0.0028728888266616394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,512,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,2048,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,256,0.005280888742870755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,64,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,32,0.0028204443968004654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,128,0.005239999956554837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,3584,0.012296000288592445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,65536,0.14766311645507812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,32,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,51200,0.11648889382680257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,6144,0.017543110582563613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,256,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,16384,0.0417431096235911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,128,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,16384,0.02164977788925171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,64,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,12288,0.03302755620744493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,1024,0.006694222076071634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,10240,0.02801777919133504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,12288,0.01793599956565433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,12288,0.02676444417900509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,65536,0.06646400027804904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,8192,0.02190133266978794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,10240,0.016185777054892648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,7168,0.020488889680968392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,51200,0.06268977456622653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,8192,0.01998222205373976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,6144,0.017844445175594754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,7168,0.013131555583741931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,7168,0.01835911141501533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,16384,0.02584444483121236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,6144,0.012069332930776807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,6144,0.01551644504070282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,4096,0.014156444205178155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,5120,0.013456000222100152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,4096,0.009708444277445475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,10240,0.02158311174975501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,51200,0.05389599998792013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,3584,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,3584,0.011767999993430244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,5120,0.015633778439627755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,5120,0.011047110789351992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,3072,0.010832000109884473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,2560,0.009687999884287516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,3072,0.008344000412358178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,4096,0.012461333639091916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,2560,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,2560,0.010031999813185798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,2048,0.0074026667409473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,2048,0.008650666309727563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,1536,0.007229333122571309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,1536,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,1024,0.005239999956554837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,65536,0.08430666393703884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,1024,0.006612444503439798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,3584,0.012086222569147745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,1024,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,768,0.006476444502671559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,512,0.003656889001528422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,768,0.005971555494599872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,3072,0.010762666662534079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,256,0.0032311110860771606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,512,0.006072888771692912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,512,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,128,0.0031831111345026228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,256,0.0052844443255000645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,64,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,768,0.0042453333735466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,8192,0.01419911119672987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,64,0.005245333330498801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,32,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,1536,0.007328000333574083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,65536,0.07381066348817614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,51200,0.09715911414888169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,128,0.005993777679072485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,128,0.004960888789759742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,16384,0.0354639987150828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,51200,0.05162577827771505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,51200,0.0564862224790785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,65536,0.12306222650739883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,12288,0.028382221857706707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,16384,0.020961778031455144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,10240,0.02496266696188185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,12288,0.017281777328915067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,12288,0.02516088883082072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,8192,0.021335111724005804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,10240,0.015346666177113851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,10240,0.01995644470055898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,7168,0.019121777680185106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,8192,0.013782222237851886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,8192,0.02029511166943444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,6144,0.01697422232892778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,7168,0.012785777449607849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,7168,0.016534222496880423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,6144,0.011748444702890186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,16384,0.02362577782736884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,6144,0.016232000456915963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,4096,0.011721777419249216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,5120,0.010715555813577441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,5120,0.01277955538696713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,3584,0.010382222632567087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,4096,0.011951111257076263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,3584,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,3072,0.010317333042621613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,5120,0.015225778023401896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,3072,0.010405333505736457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,256,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,2560,0.008990222381220924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,4096,0.00941777808798684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,2048,0.007982222570313348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,2560,0.00907466643386417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,3584,0.01108977778090371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,1536,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,1536,0.006991110742092133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,3072,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,1024,0.004937777916590373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,1536,0.006643555644485686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,2560,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,768,0.004228444563017951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,1024,0.006351111249791251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,2048,0.007281777759393056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,768,0.006355555521117316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,768,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,512,0.006267555471923616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,256,0.0031991110493739447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,512,0.005640000104904175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,256,0.005978666659858491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,128,0.0028728888266616394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,1024,0.00630044440428416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,128,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,128,0.004963555683692296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,512,0.003567111161020067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,32,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,64,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,32,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,65536,0.10681333144505818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,256,0.005247999810510212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,2048,0.0083137775460879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,51200,0.08482666810353597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,65536,0.06692355871200562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,64,0.0025964443468385274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,2048,0.007828444242477417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,16384,0.0321724447939131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,51200,0.05118933320045471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,12288,0.025729777084456548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,16384,0.020232889387342665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,16384,0.02128711177243127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,10240,0.022259554929203455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,12288,0.01663733356528812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,12288,0.024711999628278945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,8192,0.019588443968031142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,65536,0.0635075569152832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,10240,0.01923288901646932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,7168,0.01701866587003072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,8192,0.013407111167907715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,8192,0.017242666747834947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,7168,0.012423999607563019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,7168,0.017284444636768766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,6144,0.011151110960377587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,6144,0.013456888496875763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,10240,0.014928888943460254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,32,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,5120,0.012438221938080259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,4096,0.008674666285514832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,4096,0.011503111157152386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,6144,0.015575110912322998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,3584,0.008105778031879002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,3072,0.009367999931176504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,5120,0.01016088906261656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,65536,0.061353776190016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,2560,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,3072,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,51200,0.049625777535968356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,3584,0.01036444471942054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,2048,0.007634667058785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,2048,0.007299555672539606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,1536,0.005990222096443176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,5120,0.013951111170980664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,1536,0.006959999601046245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,4096,0.012080888781282636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,3072,0.00960533320903778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,1024,0.00630044440428416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,2560,0.007631999750932057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,2560,0.008374222450786168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,768,0.006314666734801398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,512,0.0035573335157500375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,512,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,512,0.006223111102978389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,256,0.003175111073586676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,768,0.003922666526503033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,256,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,128,0.002883555574549569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,3584,0.010953777366214328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,128,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,1536,0.006792000184456508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,128,0.00498577786816491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,64,0.002575111058023241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,32,0.002955555501911375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,65536,0.08600266774495442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,64,0.004960888789759742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,32,0.005286222116814719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,51200,0.06729688909318712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,256,0.0052382221652401825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,768,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,65536,0.06105866697099474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,51200,0.047777778572506376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,51200,0.04658044377962748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,1024,0.0046746668716271715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,16384,0.01962577799956004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,16384,0.019242667489581637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,10240,0.019179556104871962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,12288,0.016230222251680162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,1024,0.005992888990375731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,12288,0.020959110723601446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,8192,0.01683111157682207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,10240,0.014543111125628153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,7168,0.013815999858909182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,10240,0.01773599949147966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,65536,0.05902577771080864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,2048,0.007425777614116669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,8192,0.01719466679626041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,16384,0.027070222629441157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,7168,0.011752000285519494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,5120,0.011033777561452655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,7168,0.014748444159825643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,4096,0.01036622209681405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,6144,0.012915555801656512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,5120,0.012109333442317115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,4096,0.008348444269763099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,4096,0.010853333605660332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,3584,0.008039111064540016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,3584,0.009444444543785518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,8192,0.012822222378518848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,6144,0.012444444828563266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,2560,0.0074106670088238185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,3072,0.007648000286685095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,6144,0.010390222072601318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,2048,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,2560,0.007318221860461765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,2560,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,3584,0.008707555631796518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,2048,0.006991110742092133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,3072,0.008018666671382057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,1536,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,1024,0.003910222401221593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,1536,0.006635555376609166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,3072,0.00869688888390859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,1024,0.0063946665161185795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,768,0.0035502223504914176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,5120,0.009026666482289633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,1024,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,512,0.003253333270549774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,768,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,256,0.003212444484233856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,768,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,512,0.0052791109515561005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,128,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,256,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,128,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,64,0.0028524444335036804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,12288,0.021528888079855178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,32,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,64,0.00526577772365676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,512,0.006037333359320958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,65536,0.0781502193874783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,32,0.005250666704442766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,256,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,51200,0.05991466840108236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,65536,0.05815822548336453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,65536,0.05729689200719198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,51200,0.04684177703327603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,51200,0.044813334941864014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,16384,0.019324445062213473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,16384,0.018239999810854595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,10240,0.0166320006052653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,12288,0.015999999311235216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,12288,0.02014488809638553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,8192,0.0148062225845125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,2048,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,10240,0.01789777808719211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,7168,0.013536888692114087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,8192,0.01242844429281023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,8192,0.015473778049151102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,16384,0.023810666468408372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,7168,0.011396444506115384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,128,0.004907555464241239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,12288,0.019300444258583915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,6144,0.009923555784755284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,6144,0.01294844432009591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,4096,0.009681777821646797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,5120,0.011391999820868174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,10240,0.014332445131407844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,1536,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,3584,0.008656000097592672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,4096,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,4096,0.008912000391218398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,3072,0.0076248885856734375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,3584,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,2560,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,3584,0.009027555584907532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,3072,0.007903111477692923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,5120,0.010544000400437249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,2560,0.007702222300900354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,5120,0.00906844437122345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,2048,0.006992888947327931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,2048,0.0069857777820693115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,1536,0.006678222368160884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,1024,0.003912000192536248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,1536,0.006646222124497096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,3072,0.007638221813572778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,768,0.003527111063400904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,7168,0.013458666702111563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,2048,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,512,0.0032239999208185407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,768,0.006238222122192383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,768,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,256,0.0028524444335036804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,512,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,512,0.005285333428117964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,128,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,256,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,256,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,64,0.002507555608948072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,6144,0.011762667033407422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,32,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,1024,0.005734222216738596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,2560,0.007359111474619971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,65536,0.06330844428804186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,32,0.004949333353175057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,65536,0.05708533525466919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,51200,0.05081689026620653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,65536,0.05519022213088142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,16384,0.0206977774699529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,51200,0.04598044355710348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,51200,0.043098668257395424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,128,0.005965333431959152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,16384,0.01922044489118788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,128,0.004941333499219683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,10240,0.01669066647688548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,12288,0.015867556134859722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,64,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,12288,0.017826666434605915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,8192,0.015104000767072042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,10240,0.013949332965744866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,1536,0.004575110971927643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,7168,0.01240355521440506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,8192,0.011808888779746162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,8192,0.014125333891974555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,6144,0.011111111276679568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,7168,0.013348444468445249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,1024,0.006302222195598815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,5120,0.010022222167915767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,6144,0.009970666633711921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,6144,0.011583111352390714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,4096,0.009326222042242685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,5120,0.00904088881280687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,5120,0.009913778139485253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,3584,0.008296888735559251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,4096,0.0086986662613021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,7168,0.010381333529949188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,3072,0.007281777759393056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,16384,0.017455110947291057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,2560,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,3072,0.007660444411966536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,3072,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,2048,0.005335999859703912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,2560,0.007048000064161088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,4096,0.00833155545923445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,1536,0.004612444589535396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,2048,0.006299555715587404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,3584,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,12288,0.017535110314687092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,1024,0.005586666779385672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,2560,0.007341333561473423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,768,0.003903111235962974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,1024,0.006259555617968242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,10240,0.015073777900801765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,512,0.0032311110860771606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,768,0.006041777630647023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,768,0.0053119998839166425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,256,0.00300177786913183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,1536,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,512,0.0053004444473319584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,128,0.0028924445311228433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,256,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,256,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,128,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,128,0.004960000101062986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,32,0.002554666664865282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,64,0.00518488883972168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,2048,0.007024889190991719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,32,0.004962666581074397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,3584,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,1024,0.0038773334688610504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,51200,0.045055111249287925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,65536,0.0564791096581353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,16384,0.01869511107603709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,51200,0.0451822214656406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,51200,0.0414737794134352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,12288,0.015786666009161208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,16384,0.018784888916545443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,16384,0.017264889346228707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,10240,0.014123555686738757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,12288,0.015493333339691162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,65536,0.05759555763668484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,12288,0.01629955569903056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,8192,0.012698666916953193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,10240,0.013442666994200813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,10240,0.01440177857875824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,1536,0.006049777898523543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,8192,0.013106666505336761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,6144,0.01088711122671763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,7168,0.01055555542310079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,64,0.0028657778683635923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,5120,0.010136889086829292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,6144,0.009707555174827576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,6144,0.010019555687904358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,4096,0.008835555778609382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,5120,0.008683555656009251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,4096,0.00832799987660514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,4096,0.007670222057236566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,7168,0.011759999725553723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,8192,0.011750222080283694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,3072,0.006948444578382704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,3584,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,3584,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,7168,0.01181422256761127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,2560,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,3072,0.006959111326270633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,2048,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,5120,0.009128000173303816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,2560,0.006665777828958299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,1536,0.004596444467703502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,2048,0.007068444457319047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,2048,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,512,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,3584,0.007978666987684038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,1536,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,65536,0.05286133289337158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,768,0.003576000117593341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,1024,0.005592888842026393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,3072,0.007631111476156447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,512,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,768,0.005399999933110342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,2560,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,256,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,512,0.005988444305128521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,512,0.005274666680230035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,128,0.002883555574549569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,256,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,1024,0.0038977778620190094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,64,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,128,0.005643555687533484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,32,0.002558222247494592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,64,0.004956444518433677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,65536,0.05064622229999966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,32,0.004974222017659081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,65536,0.0555937753783332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,65536,0.0506951113541921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,51200,0.0368115570810106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,256,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,1536,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,51200,0.04503555430306328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,51200,0.04006400042110019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,12288,0.014975999792416891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,16384,0.016731555263201397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,12288,0.014655111564530266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,10240,0.013207111093733044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,16384,0.016525333126386006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,10240,0.013342222405804528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,1024,0.006301333506902059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,8192,0.01074488874938753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,16384,0.01867644488811493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,12288,0.0148062225845125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,7168,0.010152888794740042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,6144,0.010647999743620554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,8192,0.012806221842765808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,128,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,5120,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,6144,0.009569777382744683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,6144,0.008779555559158325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,8192,0.012805333567990197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,5120,0.009007111191749573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,5120,0.008343111309740279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,4096,0.008325333396593729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,7168,0.010121777653694153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,3584,0.008014221986134848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,3584,0.007305777735180325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,3072,0.006311111152172089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,10240,0.012439111040698158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,3072,0.007941332956155142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,2560,0.006028444402747684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,3072,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,2048,0.0052391112678580815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,4096,0.00832177781396442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,2560,0.006675555474228329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,2048,0.0069520001610120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,3584,0.00757155567407608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,2048,0.006085333310895496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,4096,0.00756888919406467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,1024,0.0038773334688610504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,1536,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,768,0.0037173334923055437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,1536,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,1024,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,512,0.0035457776652442086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,768,0.006132444573773279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,768,0.0052791109515561005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,512,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,512,0.005218666460778978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,256,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,256,0.005032888717121548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,128,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,1024,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,64,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,128,0.005639111002286275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,32,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,7168,0.011047110789351992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,65536,0.036362667878468834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,1536,0.004319111092223061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,32,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,51200,0.02864444586965773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,65536,0.055105778906080455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,65536,0.0484444432788425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,16384,0.013859555953078799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,51200,0.04451644420623779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,51200,0.03875822159979079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,12288,0.012477333347002665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,16384,0.017252445220947266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,16384,0.016200888488027785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,10240,0.011392888923486074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,12288,0.013461333182122974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,12288,0.014193778236707052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,8192,0.010427555276287926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,10240,0.012160888976520963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,10240,0.011866666376590729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,7168,0.009755555954244401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,8192,0.010050666828950247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,7168,0.010217777556843227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,7168,0.009375111096435124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,6144,0.009707555174827576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,256,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,768,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,5120,0.008971555365456475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,64,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,6144,0.008683555656009251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,4096,0.007631999750932057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,5120,0.009008888569143083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,5120,0.008201777935028076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,3584,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,4096,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,3584,0.007671111159854465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,3584,0.007327111230956183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,3072,0.00638133328821924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,2560,0.0073520003093613525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,3072,0.007670222057236566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,6144,0.009431111315886179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,3072,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,128,0.0026337777574857077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,2560,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,2560,0.00647199981742435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,1536,0.004580444345871608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,2048,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,2048,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,1024,0.0038764443662431505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,1536,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,1536,0.005968889014588461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,768,0.00360000009338061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,1024,0.006270222365856171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,1024,0.005584888988071018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,512,0.003198222153716617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,768,0.006010666489601135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,768,0.0052826665341854095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,2048,0.004910222358173794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,256,0.002934222213096089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,512,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,128,0.0028693332440323303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,256,0.005625777774386936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,64,0.002528888897763358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,256,0.00490399988161193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,128,0.00564533347884814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,32,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,128,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,64,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,32,0.004907555464241239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,51200,0.02183733383814494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,65536,0.046847111648983426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,512,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,16384,0.012090666426552666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,8192,0.01091466678513421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,4096,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,12288,0.011404444773991903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,16384,0.016051555673281353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,16384,0.015503999259736804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,65536,0.026093333959579468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,10240,0.01036533299419615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,12288,0.012473777764373355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,8192,0.009288888838556077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,2560,0.005643555687533484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,65536,0.05442755421002706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,7168,0.009696889254781935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,51200,0.04382755690150791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,51200,0.03769066598680284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,6144,0.0088746663596895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,7168,0.009369778136412302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,12288,0.01348000019788742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,5120,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,10240,0.012070222033394707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,10240,0.011349333657158745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,4096,0.007650666766696506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,5120,0.008071111308203803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,8192,0.009746666583749983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,3584,0.0069448889957533936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,4096,0.008360889222886827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,7168,0.010255110760529837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,3072,0.006254222244024277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,3584,0.008015111088752747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,3584,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,2560,0.005586666779385672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,3072,0.0069884442620807225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,2560,0.007290666302045186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,2560,0.0064791109826829694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,5120,0.008952000074916417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,8192,0.010969777901967367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,1536,0.005630222045713001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,2048,0.0069955554273393415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,2048,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,1024,0.004239999999602635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,1536,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,1536,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,768,0.003926222109132343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,6144,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,1024,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,512,0.003532444437344869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,768,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,768,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,2048,0.006607999818192587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,512,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,4096,0.007538666327794393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,256,0.004902222090297275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,128,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,1024,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,6144,0.008665777742862701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,64,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,128,0.004899555610285865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,32,0.0028524444335036804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,256,0.0031742221779293488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,64,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,768,65536,0.023730667101012334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,256,0.005721777677536011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,65536,0.054323554039001465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,768,51200,0.019121777680185106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,3072,0.007682667010360294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,768,16384,0.01126488877667321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,51200,0.03742488887574937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,16384,0.016062221593326993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,512,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,16384,0.014496889379289416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,32,0.00488800017370118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,768,10240,0.009690666364298927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,12288,0.012110222544935016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,10240,0.011075555450386472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,51200,0.04352000024583605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,128,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,8192,0.01075111081202825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,8192,0.009738666315873465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,768,7168,0.009363555245929295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,768,6144,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,7168,0.010032888915803697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,12288,0.013447999954223633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,768,5120,0.007970666719807519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,65536,0.045928888850741915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,6144,0.008739555875460306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,768,4096,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,5120,0.00904622260067198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,768,8192,0.008658666577604082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,5120,0.008225777910815345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,768,3584,0.0069511110583941145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,4096,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,768,3072,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,4096,0.007426666716734569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,3584,0.00798044436507755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,768,2560,0.005631111148330901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,3584,0.007293333609898885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,3072,0.007663999994595845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,768,2048,0.005210666606823604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,2560,0.007342221836249034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,7168,0.009353777600659264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,6144,0.009349333743254343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,768,1536,0.005285333428117964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,2048,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,10240,0.01180355581972334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,1536,0.006656000183688269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,1536,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,768,768,0.0038924444880750445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,1024,0.005618666609128316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,768,12288,0.010949333508809408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,768,0.005967999911970562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,3072,0.006990222467316522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,2560,0.00657066661450598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,512,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,768,256,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,768,1024,0.0042160000238153665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,768,128,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,256,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,1024,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,768,64,0.002528000002106031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,128,0.004910222358173794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,768,32,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,768,512,0.0032355555643637977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,64,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,32,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,512,0.005295111073387994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,256,0.005660444498062134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,512,51200,0.016171556380059984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,65536,0.05041066805521647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,512,16384,0.010423111418883005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,128,0.005649777750174205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,51200,0.03664711117744446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,512,12288,0.010093332992659675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,16384,0.01593955523437924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,16384,0.014480888843536377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,512,65536,0.017867555220921833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,512,10240,0.009011555876996782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,12288,0.013456888496875763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,12288,0.012103111379676394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,2048,0.0070017774899800616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,10240,0.012104888757069906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,512,7168,0.008692444198661381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,10240,0.01110222190618515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,8192,0.010025777750545079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,65536,0.0454364452097151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,512,6144,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,7168,0.00923911150958803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,512,5120,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,6144,0.00867733359336853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,768,0.005241777747869492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,512,4096,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,5120,0.009015999734401703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,5120,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,512,8192,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,512,3584,0.006786666810512543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,4096,0.008373333348168267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,8192,0.01075644459989336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,512,3072,0.006606222026877933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,3584,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,51200,0.038698666625552706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,6144,0.009692444569534725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,3072,0.006680888848172293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,2560,0.007283555964628856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,512,2048,0.005964444329341252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,2560,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,2048,0.0069520001610120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,512,1536,0.004562666846646203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,2048,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,3584,0.007857778006129796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,512,1024,0.003912888881233004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,1536,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,1536,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,512,2560,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,512,768,0.0035333335399627686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,1024,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,512,512,0.0033902221669753394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,768,0.006260444306664997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,768,0.0052373334765434265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,512,256,0.002982222164670626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,4096,0.007329777710967594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,256,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,256,0.00490399988161193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,3072,0.007640889121426477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,512,64,0.002518222149875429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,128,0.005608888963858287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,512,32,0.0028497777465316984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,64,0.004936000125275718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,256,65536,0.017130666308932833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,512,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,65536,0.04689866635534498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,65536,0.04468889037768046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,1024,0.0062977779242727495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,128,0.004699555536111196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,256,16384,0.009153777526484596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,512,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,32,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,16384,0.014171555638313293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,256,12288,0.008031110796663497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,7168,0.010374222364690568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,256,51200,0.014800000521871777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,12288,0.012084444363911947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,10240,0.011769777370823754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,256,8192,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,51200,0.037392887804243304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,8192,0.01068800025516086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,256,7168,0.006641777853171031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,8192,0.009724444813198512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,16384,0.015825778245925903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,256,6144,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,7168,0.010046222143703038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,12288,0.01346933344999949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,7168,0.009400889277458191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,256,5120,0.006640000061856375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,6144,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,256,4096,0.006313777632183499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,6144,0.008841777841250101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,5120,0.009003555609120263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,256,3584,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,5120,0.008046222229798635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,51200,0.03485511077774896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,256,3072,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,4096,0.0069875551594628235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,3584,0.007978666987684038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,256,2560,0.006329777754015393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,3584,0.007318221860461765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,3072,0.007322666545708974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,256,2048,0.005591111050711737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,3072,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,2560,0.007300444775157505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,256,1536,0.004565333326657613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,512,128,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,2048,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,256,10240,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,256,1024,0.003900444342030419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,2048,0.005984000033802456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,1536,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,256,768,0.0035573335157500375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,1536,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,256,512,0.003237333355678452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,768,0.005963555640644497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,768,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,512,0.005618666609128316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,512,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,256,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,256,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,2560,0.006331555545330048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,128,0.005631999837027655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,256,64,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,128,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,256,32,0.002534222271707323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,4096,0.008343111309740279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,128,65536,0.015032889114485847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,64,0.004934222333961063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,256,256,0.0029111111329661477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,32,0.0047768888374169665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,128,51200,0.012432888978057437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,65536,0.04607377780808342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,65536,0.04228977693451775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,256,128,0.002578666640652551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,51200,0.03735555542839898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,51200,0.03367822368939718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,128,12288,0.007383999725182851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,16384,0.01588355501492818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,16384,0.014488889111412896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,10240,0.011065777805116443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,128,10240,0.007303111255168915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,12288,0.012109333442317115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,128,8192,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,10240,0.012075555821259817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,10240,0.011077333655622272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,128,7168,0.006310222049554189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,8192,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,7168,0.009024000002278222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,1024,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,128,6144,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,1024,0.005583111196756363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,128,5120,0.006007110906971826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,6144,0.008354666332403818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,8192,0.01070755554570092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,128,4096,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,7168,0.010027555955780877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,4096,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,4096,0.007631111476156447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,6144,0.009357333183288574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,128,3072,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,5120,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,3072,0.006952889263629913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,128,16384,0.008361777497662438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,3072,0.007700444095664554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,2560,0.007299555672539606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,128,2048,0.005663110978073544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,12288,0.01314311143424776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,3584,0.007864000068770515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,128,1536,0.004592888885074192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,2048,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,5120,0.008723555339707268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,1536,0.006671111202902264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,128,2560,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,1024,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,128,3584,0.0061919999619325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,1024,0.00573600000805325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,2560,0.006406222366624408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,128,512,0.0032257777121331957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,768,0.005041777673694822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,512,0.005768888940413793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,512,0.005320888840489917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,128,256,0.0028897778441508612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,128,1024,0.003903999924659729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,1536,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,128,128,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,256,0.004957333207130432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,128,64,0.002543111021320025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,128,0.005663110978073544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,128,0.004908444566859139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,768,0.006072000082996156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,64,65536,0.01145866678820716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,2048,0.00674933319290479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,32,0.00490488898422983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,65536,0.042248000701268516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,51200,0.03394044438997904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,256,0.005631999837027655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,64,12288,0.00664444433318244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,128,768,0.0033679999825027254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,64,10240,0.0070053330726093715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,16384,0.014495111174053617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,64,8192,0.0070168889231152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,12288,0.011736888852384357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,3584,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,64,7168,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,10240,0.01088177743885252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,64,6144,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,64,5120,0.005639111002286275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,7168,0.009023110899660323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,64,4096,0.005591111050711737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,5120,0.007972444097201029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,64,3584,0.005895111295912001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,128,32,0.002513777878549364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,64,3072,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,3584,0.006822222222884496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,64,2560,0.005611555443869696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,64,51200,0.010374222364690568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,64,2048,0.005590222362014983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,8192,0.009694221946928237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,64,1536,0.004262222184075249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,2560,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,6144,0.008678221868144141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,2048,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,64,768,0.0035626664757728577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,4096,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,64,512,0.0032088889016045462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,768,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,64,0.004957333207130432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,3072,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,64,16384,0.0070017774899800616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,256,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,1536,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,64,64,0.0025048889219760895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,1024,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,128,0.004879111217127906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,64,256,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,32,65536,0.01108088923825158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,512,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,32,51200,0.010395555860466426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,65536,0.04152444336149428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,32,16384,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,51200,0.033943110042148165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,32,12288,0.0069520001610120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,64,32,0.002521777732504739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,64,0.00490399988161193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,32,10240,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,32,0.004648888690604104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,10240,0.011053333679835001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,32,7168,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,64,128,0.002578666640652551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,32,6144,0.006000888844331105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,7168,0.009322666459613377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,32,5120,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,16384,0.014138667119873894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,32,4096,0.00563822231358952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,6144,0.008372444245550368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,32,3584,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,32,3072,0.005912000106440649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,5120,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,32,2560,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,4096,0.007327111230956183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,32,2048,0.0052373334765434265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,3072,0.006979555719428592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,32,1536,0.00424888895617591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,2560,0.006334222025341458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,32,1024,0.0038773334688610504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,2048,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,32,768,0.0033644443998734155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,1536,0.005983110931184556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,32,512,0.0032311110860771606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,768,0.005230222311284807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,32,256,0.0028773333049482773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,32,8192,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,512,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,3584,0.007142222589916653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,32,64,0.0025617778301239014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,64,1024,0.003567999849716822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,32,32,0.0025395556456512874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,8192,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,128,0.004969777746333016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,64,0.004944889081848992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,32,0.004605333424276776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,1024,0.00562755556570159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,12288,0.012102222277058495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,51200,0.48682843314276797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,9,32,128,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,16384,0.34256622526380753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,51200,1.02256531185574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,16384,0.162209775712755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,12288,0.24608177608913848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,16384,0.13509599367777506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,12288,0.14061244328816733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,10240,0.20605244901445177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,12288,0.11413244406382243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,256,0.004613333278232151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,10240,0.1250213384628296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,8192,0.15447555647956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,12288,0.12713600529564753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,16384,0.17007377412584093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,8192,0.10739733113182916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,7168,0.13321067227257624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,51200,0.4334835476345486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,8192,0.07123821973800659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,10240,0.10861955748664008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,7168,0.09461777740054661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,8192,0.09114133649402195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,7168,0.06321955389446683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,10240,0.09011289146211411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,6144,0.08223199844360352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,5120,0.09674844476911758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,6144,0.05475733346409268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,4096,0.08100710974799262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,5120,0.07292089197370741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,6144,0.06963733169767591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,5120,0.0600222216712104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,4096,0.039444443252351545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,4096,0.04973511232270134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,3584,0.05447377761205038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,7168,0.07996977700127496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,5120,0.04718044400215149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,4096,0.06024266613854302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,6144,0.11485600471496582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,51200,0.49596532185872394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,2560,0.050359110037485756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,3072,0.04032533367474874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,3584,0.03529422150717841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,2560,0.03502488798565335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,2560,0.04078844520780776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,2048,0.04177688890033298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,2560,0.02749155627356635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,3072,0.04814666509628296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,2048,0.034049779176712036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,3072,0.031256887647840716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,1536,0.03256355391608344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,2048,0.03120533294147915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,1536,0.019256000717480976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,3072,0.05958133273654514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,1024,0.023392000132136877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,3584,0.06877866718504164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,1024,0.021350221501456365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,768,0.01886311173439026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,768,0.01616533266173469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,1536,0.02608088817861345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,1536,0.025850666893853083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,512,0.013421333498424955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,1024,0.019902222686343722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,512,0.012630222572220696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,2048,0.02403555479314592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,256,0.008223111430803934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,512,0.01293511109219657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,3584,0.045371555619769625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,128,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,256,0.010748444332016839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,768,0.015092444088723926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,768,0.0181040002240075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,128,0.008001777860853408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,64,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,32,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,1024,0.017309332887331646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,64,0.009467555416954888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,32,0.00999466660949919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,256,0.010013333625263637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,256,0.012093333734406365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,128,0.010727110836240979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,65536,0.9717742072211372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,128,0.010832889212502373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,512,0.01498311095767551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,51200,0.7489368650648328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,51200,0.3887733353508844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,16384,0.2530968983968099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,65536,0.43292978074815536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,65536,0.5139742427402073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,16384,0.12229155169592963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,12288,0.19359732998741996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,51200,0.36440356572469074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,10240,0.16305243968963623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,12288,0.08534400330649482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,12288,0.13696889082590738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,16384,0.13526933723025852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,10240,0.0739484429359436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,16384,0.1301342248916626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,8192,0.14012711577945286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,12288,0.10214577780829535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,8192,0.05758133199479845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,7168,0.12281421820322673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,8192,0.07504622141520183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,7168,0.09079911311467488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,10240,0.09946044286092122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,6144,0.10614222288131714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,7168,0.05214488837454054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,7168,0.06472355789608426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,6144,0.04532000091340807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,5120,0.08970489104588826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,10240,0.08831555313534206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,6144,0.05662577682071262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,5120,0.05923910935719808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,5120,0.03920711080233256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,5120,0.049950222174326576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,4096,0.048523555199305214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,51200,0.3884897761874729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,4096,0.04213066564665901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,65536,0.5128444565667046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,3584,0.04494400156868828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,3584,0.029430223835839167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,3072,0.056107554170820445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,3584,0.03817066550254822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,4096,0.07308000326156616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,2560,0.04721511072582669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,3072,0.0313991109530131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,3072,0.03982399900754293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,3584,0.06358133422003852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,2560,0.023014222582181294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,4096,0.03338844577471415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,8192,0.10644533236821492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,2048,0.02847822176085578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,1536,0.029576000240114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,3072,0.03472800056139628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,1536,0.022258667482270136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,2560,0.0332951115237342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,1536,0.01648000048266517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,1536,0.023344889283180237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,1024,0.016516443755891588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,2048,0.019824888971116807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,768,0.01672266589270698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,1024,0.01791288952032725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,768,0.013984888792037964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,768,0.015488889482286243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,1024,0.021593777669800654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,768,0.013544888959990608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,512,0.011359111302428775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,1024,0.015207111835479736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,512,0.011284444067213269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,512,0.014480888843536377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,256,0.007258666886223688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,6144,0.06807377603318956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,256,0.010878221856223213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,128,0.0038693332009845306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,2048,0.027088888817363318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,128,0.0073520003093613525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,128,0.01035911093155543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,64,0.0035217776894569397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,128,0.009689778089523315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,32,0.003569777641031477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,512,0.011053333679835001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,64,0.00870222184393141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,256,0.00871288859181934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,32,0.00903200027015474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,65536,0.32408534155951607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,51200,0.25119733810424805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,65536,0.17329155074225533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,65536,0.16460088888804117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,16384,0.08755733569463094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,51200,0.12279644277360703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,256,0.009729777773221334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,16384,0.049606220589743726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,16384,0.052517334620157875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,12288,0.06744177473915948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,51200,0.13036978244781494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,51200,0.14065689510769314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,2560,0.030527111556794908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,10240,0.05485688977771335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,12288,0.051113777690463595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,12288,0.041238221857282854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,10240,0.04127022292878892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,8192,0.04296888907750448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,10240,0.028749333487616643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,10240,0.036299556493759155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,7168,0.037011557155185275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,8192,0.02441066669093238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,65536,0.1770168940226237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,8192,0.03146222233772278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,2048,0.03786133395300971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,6144,0.032664888434939914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,7168,0.03210044569439358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,7168,0.028225776222017076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,6144,0.01992266707950168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,5120,0.02821777926550971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,6144,0.025780444343884785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,5120,0.02539377742343479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,5120,0.017671111557218764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,4096,0.02367111047108968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,5120,0.023364444573720295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,7168,0.022218666142887537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,3584,0.02050933407412635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,12288,0.036665777365366616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,6144,0.027843554814656574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,3584,0.019349333312776353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,3072,0.018208889497650992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,3584,0.01883555617597368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,3072,0.017575111654069688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,4096,0.015847111741701763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,3072,0.01718133356836107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,4096,0.019875556230545044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,3072,0.013432888521088494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,16384,0.04775466521581014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,2048,0.013031110995345645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,2560,0.015257777439223396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,2560,0.01590222285853492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,8192,0.039727999104393855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,2048,0.01390044391155243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,1536,0.010924444430404238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,2048,0.01475822263293796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,4096,0.021143111917707656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,1536,0.011333333121405708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,1024,0.008403555386596257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,1536,0.013020444247457715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,2560,0.012346666720178394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,768,0.0069546666410234236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,1024,0.011613333390818702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,2048,0.011042666931947073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,768,0.008037333687146505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,512,0.005220444252093633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,768,0.008343111309740279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,1536,0.009703111317422655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,2560,0.015847999188635085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,256,0.0035253332720862497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,512,0.00666222224632899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,512,0.009033777647548253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,1024,0.009016888837019602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,128,0.002895999906791581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,256,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,256,0.00906488878859414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,128,0.005608888963858287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,128,0.0069395556218094295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,128,0.009008888569143083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,32,0.0029093333416514923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,64,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,32,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,1024,0.009047110875447592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,65536,0.2471004327138265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,65536,0.151055998272366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,64,0.0028888889484935333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,768,0.010187555518415239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,51200,0.1907635529836019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,512,0.00758844448460473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,65536,0.13974400361378989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,51200,0.11226044760810004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,16384,0.06667111317316692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,3584,0.014523555835088095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,51200,0.10155200295978123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,12288,0.0518817769156562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,16384,0.03775822122891744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,65536,0.13397955894470215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,16384,0.041366222831938006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,10240,0.044165333112080894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,12288,0.029626667499542236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,51200,0.10258755418989395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,12288,0.0325528879960378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,8192,0.03719733489884271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,10240,0.02599999970859951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,10240,0.04070399867163764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,7168,0.032384001546435885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,256,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,8192,0.025855110751258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,16384,0.04125600059827169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,7168,0.030215111043718126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,7168,0.020184000333150227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,7168,0.023228444986873206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,12288,0.04491110973887973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,6144,0.0181040002240075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,6144,0.024822221861945257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,6144,0.021321778496106465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,5120,0.016350222958458793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,8192,0.021998221675554912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,4096,0.022615111536449854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,6144,0.028808888461854722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,4096,0.019408888287014432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,3584,0.01953155630164676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,4096,0.01703377730316586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,4096,0.014257777068350049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,5120,0.02535733415020837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,3584,0.01331822243001726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,3584,0.017659554878870647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,3584,0.015844444433848064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,10240,0.02906399965286255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,8192,0.030573334958818223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,3072,0.014487110906177096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,3072,0.01610577768749661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,3072,0.015314666761292351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,2048,0.013096888860066732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,2560,0.011306666665607028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,2560,0.013839999834696451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,2560,0.014149333039919535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,2048,0.010202666951550378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,2048,0.012768000364303589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,2048,0.013575110998418597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,3072,0.01722133325205909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,1024,0.007228444847795699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,1536,0.010703999963071613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,5120,0.022422222627533808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,2560,0.01551822159025404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,768,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,1024,0.00830488900343577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,1024,0.010382222632567087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,768,0.007807999849319458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,1536,0.008360000120268928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,512,0.006642666541867786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,1536,0.012239110966523489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,5120,0.01910044418440925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,256,0.0032435556252797446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,512,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,512,0.008581333690219456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,256,0.006682666639486949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,512,0.003875555677546395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,1536,0.008743111458089616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,1024,0.00815288888083564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,128,0.008993777963850234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,128,0.0069360000391801195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,64,0.0029253332565228143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,32,0.0028942223224374983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,256,0.005786666853560342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,128,0.0028568889117903183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,64,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,65536,0.2065804402033488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,32,0.005980444451173146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,128,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,65536,0.1403599977493286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,65536,0.13287556171417236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,51200,0.10726310809453328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,16384,0.05899733304977417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,768,0.007383999725182851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,51200,0.1024071110619439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,65536,0.11799821588728164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,16384,0.03937777876853943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,16384,0.03603466682963901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,16384,0.03776266508632236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,51200,0.16251556078592935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,10240,0.03827733463711209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,12288,0.03565866748491923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,12288,0.029655112160576716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,10240,0.03121333320935567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,256,0.008998221821255153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,10240,0.026365333133273657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,10240,0.02477688921822442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,51200,0.08972177902857463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,8192,0.027710222535663184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,8192,0.0207368897067176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,7168,0.028991109795040552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,12288,0.02850133180618286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,7168,0.01935822268327077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,6144,0.0258568889564938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,7168,0.02122666603989071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,8192,0.03166044420666165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,12288,0.044512000348832875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,6144,0.01790577835506863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,5120,0.022368000613318548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,6144,0.019488889310095046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,8192,0.023366222778956097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,5120,0.01812355551454756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,5120,0.01738133364253574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,7168,0.024142222272025213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,768,0.009367999931176504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,4096,0.01613066593805949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,6144,0.02146311104297638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,3584,0.014709333578745524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,3072,0.01594666639963786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,3584,0.013136000268989138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,4096,0.019242667489581637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,3072,0.013405332962671915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,2560,0.014380445082982382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,3072,0.014834667245546976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,2560,0.01222400036123064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,2560,0.01369866645998425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,3584,0.015728889240158927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,5120,0.015782222151756287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,2048,0.011045333411958484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,3072,0.012875555290116204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,1536,0.008960888617568547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,3584,0.017471111483044095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,1536,0.009367999931176504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,1536,0.011735999749766456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,2560,0.010504888991514841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,1536,0.008265777594513362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,1024,0.007391110890441471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,1024,0.009157333109113906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,2048,0.009572444690598382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,4096,0.014110222458839417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,1024,0.007867555651399825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,768,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,768,0.006671999891599019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,768,0.00925066653225157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,2048,0.010780444575680627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,1024,0.006163555714819167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,256,0.003220444545149803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,512,0.007017778025733099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,512,0.00867733359336853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,256,0.00564177789621883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,256,0.006674666785531574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,4096,0.016314667132165697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,128,0.00526577772365676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,128,0.009027555584907532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,64,0.0029146667155954572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,512,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,32,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,64,0.005640888793600931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,768,0.00489688871635331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,2048,0.013181333740552267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,32,0.005768888940413793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,512,0.003536000019974179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,65536,0.087391111585829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,65536,0.07173511054780748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,51200,0.13050666120317247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,51200,0.06877155436409844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,51200,0.056687997447119824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,16384,0.04547022117508782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,65536,0.10145422485139634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,128,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,256,0.00903911143541336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,65536,0.174180441432529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,51200,0.0807795524597168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,16384,0.02721244427892897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,10240,0.0277786652247111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,16384,0.03279111120435927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,12288,0.026942221654786006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,128,0.006721777634488211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,8192,0.023694222172101338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,10240,0.02459555533197191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,8192,0.020598222812016804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,12288,0.03307022319899665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,8192,0.0148053334818946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,16384,0.023444443941116333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,12288,0.01904088921017117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,7168,0.018911111685964797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,12288,0.02757866680622101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,6144,0.018220444520314533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,10240,0.023990222149425085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,6144,0.012780444489585029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,6144,0.016935111747847665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,6144,0.018400889303949144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,8192,0.02163377735349867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,5120,0.015160888433456421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,5120,0.011622221933470832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,7168,0.013570666313171387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,7168,0.020068445139461093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,4096,0.013450666434235044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,4096,0.015490666031837463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,3584,0.011682666838169098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,7168,0.020616888999938965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,3584,0.01205244412024816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,3584,0.014705777168273926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,4096,0.0136817776494556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,3072,0.011139555937714048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,5120,0.015790222419632804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,3072,0.011386666860845355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,3072,0.013798221945762634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,4096,0.010649777948856354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,2560,0.00833422193924586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,2048,0.008359111017651027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,2560,0.013106666505336761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,5120,0.016696888539526198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,10240,0.01645155582163069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,1536,0.007052444749408298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,2048,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,2048,0.01240799989965227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,2560,0.009711110757456886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,1536,0.007019555403126611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,1024,0.005583111196756363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,1536,0.011056000159846412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,2560,0.010395555860466426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,1024,0.006676444576846228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,1024,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,1024,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,3584,0.010029333333174387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,768,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,768,0.009010666774378883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,512,0.005776888794369168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,1536,0.007686222592989604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,256,0.0031884445084465873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,512,0.008445333275530074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,512,0.006576889091067844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,768,0.004223111189073986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,256,0.006235555642180973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,256,0.0053048887186580235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,256,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,512,0.003363555504216088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,128,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,128,0.00904622260067198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,32,0.002576888849337896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,3072,0.009161777794361115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,65536,0.14997600184546575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,128,0.0028391112056043413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,768,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,64,0.002552888873550627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,51200,0.11811289522382949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,65536,0.06580533583958943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,64,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,65536,0.09943199819988674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,32,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,16384,0.041631112496058144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,51200,0.05360355642106798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,16384,0.021651556094487507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,16384,0.025895110434956018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,12288,0.033193777004877724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,16384,0.03274666600757175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,12288,0.018111111389266122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,10240,0.028582221931881372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,12288,0.02702399922741784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,51200,0.06255111429426405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,10240,0.02276977731121911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,8192,0.02431022293037838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,10240,0.024285333024130926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,10240,0.01588444411754608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,51200,0.0799786647160848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,2048,0.009272889130645329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,7168,0.018844443890783522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,8192,0.019374221563339233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,8192,0.021321778496106465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,12288,0.025546666648652818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,6144,0.016590222716331482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,7168,0.01960000064637926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,6144,0.01573244399494595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,6144,0.012304888831244575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,6144,0.018386666973431904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,65536,0.08218222194247775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,128,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,5120,0.01640888883007897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,5120,0.011351111034552256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,7168,0.013110222087966071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,4096,0.012642666697502136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,4096,0.009962666365835402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,3584,0.011069333387745751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,4096,0.01550044450494978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,5120,0.014449778530332776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,3072,0.010120888551076254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,3584,0.011569778124491373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,3584,0.014415999253590902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,5120,0.014092443717850579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,8192,0.014162666267818876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,3072,0.008445333275530074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,3072,0.014129777749379476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,7168,0.018206222189797294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,2560,0.010031110710567897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,2048,0.007733333441946242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,2560,0.013034666577974955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,3584,0.00905688852071762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,2048,0.008400888906584846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,1536,0.006614222294754452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,2560,0.009170666337013245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,4096,0.013163555827405719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,1536,0.007326222128338284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,1024,0.004572444491916233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,1536,0.01070844464831882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,2560,0.007982222570313348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,1024,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,1024,0.006351999938488007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,768,0.0038817777401871155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,1024,0.009328000247478485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,2048,0.012439111040698158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,768,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,512,0.0032995556377702286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,3072,0.010749333434634738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,768,0.009293333523803288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,256,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,1536,0.006980444822046492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,512,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,512,0.008729777402347988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,128,0.0028417776856157514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,256,0.005962666538026597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,256,0.00866844422287411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,64,0.002559111143151919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,128,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,128,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,2048,0.007328888608349695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,768,0.005971555494599872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,64,0.005274666680230035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,32,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,512,0.006121777825885349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,65536,0.07456533114115398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,51200,0.09678755866156684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,256,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,51200,0.05624800258212619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,16384,0.035558223724365234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,128,0.007991111112965478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,51200,0.07962933513853285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,16384,0.023430221610599097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,16384,0.02089777754412757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,16384,0.03244266576237149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,65536,0.06340266598595513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,12288,0.0236124445994695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,12288,0.026687999566396076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,51200,0.051679111189312406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,32,0.002560888934466574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,10240,0.020600888464185927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,8192,0.021572444174024794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,65536,0.12249599562750922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,12288,0.028488000233968098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,8192,0.013630222115251752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,7168,0.01930488811598884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,8192,0.02162577708562215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,12288,0.01732711162832048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,65536,0.09870577520794338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,7168,0.017343110508388944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,7168,0.012774222426944308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,10240,0.015832000308566622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,10240,0.023697778582572937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,5120,0.015290665957662793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,6144,0.018193778064515855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,8192,0.017877333694034152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,5120,0.013566222455766467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,4096,0.011540444360838996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,5120,0.016937777400016785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,6144,0.016918222109476726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,4096,0.012078222301271228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,3584,0.010374222364690568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,4096,0.015180443723996481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,6144,0.014833778142929077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,3584,0.011342222491900126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,3072,0.009351111120647853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,3584,0.014610666367742749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,5120,0.0107360002067354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,10240,0.02443733314673106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,2560,0.008381333616044786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,3072,0.010500444306267632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,3072,0.013783111340469785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,4096,0.009484444227483537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,2048,0.007282666862010956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,3584,0.008288000192907121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,2048,0.008204444415039485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,2048,0.007347555624114142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,2048,0.011051555474599203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,1536,0.006274666637182236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,2560,0.007638221813572778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,2560,0.012987555729018318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,1024,0.004498666773239772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,1536,0.006895111252864202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,1024,0.006362666686375936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,1024,0.009359111388524374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,768,0.003576888806290097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,2560,0.009480000370078618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,7168,0.019724445210562814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,768,0.005982222242487802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,768,0.009071110851234859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,512,0.003600888782077365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,1024,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,512,0.005674666828579373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,1536,0.007073777417341868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,512,0.008648888932334052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,512,0.006039999839332368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,256,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,768,0.006305777778228124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,256,0.008395555946562026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,3072,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,128,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,6144,0.01200977795653873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,64,0.0028737777223189673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,64,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,256,0.005912000106440649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,65536,0.10626311434639825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,32,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,1536,0.010088889135254754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,65536,0.061208003097110324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,51200,0.08491555849711101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,128,0.0028773333049482773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,51200,0.050995555188920766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,256,0.0029333333174387612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,16384,0.032136887311935425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,65536,0.09863466686672634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,32,0.002846222163902389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,16384,0.020217777954207528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,12288,0.02590755621592204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,16384,0.03216266632080078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,128,0.007962666451931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,51200,0.049561778704325356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,12288,0.01686044368478987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,12288,0.026410667432679072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,16384,0.021115556359291077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,10240,0.01983644399378035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,8192,0.019499555230140686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,10240,0.014993778533405729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,8192,0.017415111263593037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,51200,0.07938844627804227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,8192,0.02087644404835171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,7168,0.017264000243610807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,8192,0.013437333206335703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,10240,0.02199111051029629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,6144,0.01554844445652432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,7168,0.012437333663304647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,7168,0.016882666283183627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,7168,0.019609777463806998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,5120,0.01347644461525811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,6144,0.011495999991893768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,10240,0.02381422287887997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,6144,0.017528888252046373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,4096,0.010593777729405297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,5120,0.013071999781661563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,128,0.005997333261701796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,5120,0.016007999579111736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,3584,0.00923288861910502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,4096,0.008672888908121321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,4096,0.015255111787054272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,3584,0.011056000159846412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,3584,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,3584,0.014353776971499125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,5120,0.01015822258260515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,65536,0.06685866912206014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,3072,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,12288,0.021709332863489788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,4096,0.011500444677140979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,2560,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,2048,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,3072,0.00832444429397583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,6144,0.014455111490355598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,2048,0.007300444775157505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,3072,0.010004444254769219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,1536,0.005137777576843898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,3072,0.013799111048380533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,2560,0.007967111137178209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,1024,0.003899555653333664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,1536,0.006760889043410619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,1536,0.010074666804737514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,2560,0.008311111066076491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,1024,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,1024,0.009390222529570261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,768,0.005714666512277391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,2048,0.01074577785200543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,768,0.009024000002278222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,1536,0.00665244460105896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,512,0.005648889061477449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,1024,0.00647288892004225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,256,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,512,0.008737777670224508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,2048,0.007654222349325816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,256,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,128,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,768,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,128,0.004907555464241239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,128,0.008367111285527548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,64,0.002522666628162066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,512,0.003549333247873518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,256,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,32,0.002521777732504739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,64,0.005051555732885997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,32,0.005187555733654234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,768,0.003543111185232798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,65536,0.061372445689307324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,256,0.008625778059164682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,51200,0.06691110796398587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,65536,0.09089422225952148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,2560,0.011728888584507836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,16384,0.026426666312747534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,51200,0.047600001096725464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,51200,0.0790826678276062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,16384,0.019466666711701285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,16384,0.019239111079110038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,12288,0.025921778546439275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,16384,0.031982223192850746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,12288,0.01616533266173469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,12288,0.019575999842749704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,10240,0.017893332574102614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,12288,0.026195555925369263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,10240,0.014482667048772177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,10240,0.017820444371965196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,8192,0.016196444630622864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,10240,0.023356444305843774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,8192,0.012758221891191272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,65536,0.09861599736743504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,7168,0.014145778285132514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,8192,0.02065955599149068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,7168,0.011419555379284753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,128,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,7168,0.019152000546455383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,51200,0.04654844601949056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,6144,0.013110222087966071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,5120,0.012307555311255984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,6144,0.010413332945770687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,65536,0.058872891796959766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,5120,0.012070222033394707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,5120,0.00942666663063897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,4096,0.010435555544164447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,5120,0.01647555496957567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,7168,0.015491555134455362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,512,0.00599644457300504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,4096,0.00904088881280687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,4096,0.011000889043013254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,4096,0.014905777242448596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,3072,0.009849777652157677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,3584,0.010070222119490305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,6144,0.01754399968518151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,3584,0.014031110538376702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,8192,0.01551466683546702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,3072,0.008377778033415476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,3072,0.007681777907742395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,2560,0.007329777710967594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,6144,0.012777778009573618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,2560,0.011738667057620155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,2048,0.007324444750944774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,2048,0.01072622256146537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,1536,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,2048,0.006989333364698622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,2560,0.008991999758614434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,3072,0.013188444077968597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,1536,0.006678222368160884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,1536,0.006302222195598815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,2560,0.007686222592989604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,3584,0.011398221883508893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,768,0.00423555572827657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,1024,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,1024,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,1024,0.009352888498041365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,512,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,3584,0.007984889050324758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,768,0.006299555715587404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,768,0.009001777403884465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,512,0.006018666757477655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,512,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,512,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,1024,0.004941333499219683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,128,0.002899555489420891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,256,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,2048,0.0079644446571668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,256,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,64,0.0025155554629034465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,128,0.005711110929648082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,128,0.004978666702906291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,32,0.002569777684079276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,128,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,65536,0.07972888814078437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,64,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,32,0.0052977779673205475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,51200,0.061584002441830106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,65536,0.09806933005650838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,1536,0.009993777506881291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,51200,0.044738666878806226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,16384,0.024311110377311707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,51200,0.07871111234029134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,768,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,16384,0.018636445204416912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,12288,0.01902222302224901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,65536,0.057787557442982994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,16384,0.03158400124973721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,256,0.0052871112194326185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,12288,0.019661333825853135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,12288,0.015941333439615037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,10240,0.01759733259677887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,256,0.0032097777972618737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,65536,0.05798044469621447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,10240,0.014489778214030795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,10240,0.023585778143670823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,8192,0.01515111161602868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,8192,0.012618666721714867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,8192,0.020812445216708712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,7168,0.013414222333166333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,16384,0.019290667441156175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,7168,0.014241778188281588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,7168,0.011389333340856763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,6144,0.012598222328556908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,10240,0.016536000702116225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,6144,0.017837334010336135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,6144,0.010047111246320937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,5120,0.01148888882663515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,5120,0.011762667033407422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,5120,0.009031111167536842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,8192,0.014760888285107084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,4096,0.010093332992659675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,6144,0.0129466669427024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,4096,0.014783999986118741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,3584,0.011708444191349877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,4096,0.008367111285527548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,3584,0.008346666892369589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,3584,0.008657777474986183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,3584,0.01311644415060679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,4096,0.010214221974213919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,5120,0.016520000166363187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,3072,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,3072,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,3072,0.012410666379663678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,12288,0.025750221477614507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,2560,0.007305777735180325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,2048,0.007246221933099959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,2560,0.011056000159846412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,2048,0.0070186663005087115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,1536,0.005921777751710679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,2048,0.010737777584128909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,7168,0.01889955500761668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,2560,0.008911111288600499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,1536,0.00630044440428416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,1536,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,1536,0.010053333308961656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,51200,0.046606222788492836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,1024,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,1024,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,1024,0.009080000221729279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,3072,0.010029333333174387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,512,0.0035786665976047516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,768,0.00868711123863856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,1024,0.0048942222363419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,512,0.008678221868144141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,256,0.0031928889867332247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,2560,0.007355555891990662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,768,0.004231111043029361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,256,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,256,0.00489688871635331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,2048,0.006640888750553131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,128,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,128,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,128,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,512,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,32,0.002554666664865282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,64,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,65536,0.06329244375228882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,128,0.0028497777465316984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,768,0.006267555471923616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,32,0.005247999810510212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,51200,0.05058755477269491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,65536,0.057067553202311196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,512,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,65536,0.0981902215215895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,16384,0.020584889584117465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,51200,0.04608622193336487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,51200,0.07891733116573758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,16384,0.017529777354664274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,16384,0.01921866668595208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,12288,0.017051556044154696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,768,0.005344888816277186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,16384,0.03145066565937466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,10240,0.014834667245546976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,12288,0.015837333268589444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,12288,0.01718488832314809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,12288,0.025736888249715168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,8192,0.01316177762216992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,51200,0.04301333427429199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,10240,0.022904000348514978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,8192,0.013839110732078552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,7168,0.012455110748608908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,8192,0.02083200050724877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,256,0.008040889269775814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,7168,0.01071111112833023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,65536,0.05492977632416619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,7168,0.019004444281260174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,10240,0.013957333233621387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,6144,0.011866666376590729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,8192,0.011873777541849347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,6144,0.01771111124091678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,5120,0.010760889285140567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,6144,0.009730666875839233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,5120,0.009185777770148383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,7168,0.01275466630856196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,4096,0.009697777529557547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,64,0.0028497777465316984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,4096,0.008611555728647444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,4096,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,3584,0.007991111112965478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,5120,0.010716444088353051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,5120,0.015531554818153381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,3072,0.008656000097592672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,3584,0.008167111211352879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,10240,0.015496888094478183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,3584,0.009111111362775167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,2560,0.007698666718271043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,3072,0.012243555651770698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,2560,0.007002666592597961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,2560,0.011427555647161273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,2560,0.007294221884674496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,6144,0.011002666420406766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,3584,0.01314755529165268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,2048,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,2048,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,1536,0.0059164443777667145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,3072,0.007755555212497711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,1024,0.0052764444715446895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,1536,0.00664533343580034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,1536,0.006268444574541516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,1536,0.010029333333174387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,768,0.00424177779091729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,1024,0.005611555443869696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,2048,0.006607999818192587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,1024,0.009408888717492422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,512,0.0038764443662431505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,768,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,768,0.009009777671760982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,512,0.0052782222628593445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,4096,0.014506666196717156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,256,0.0035377778112888336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,512,0.008767999708652496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,1024,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,256,0.006071110980378256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,256,0.005331555588377847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,768,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,128,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,128,0.005005333158704969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,512,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,32,0.0025457777082920074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,128,0.008471999731328752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,65536,0.05756444401211209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,32,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,65536,0.052674667702780835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,65536,0.0977333320511712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,64,0.002887111157178879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,51200,0.045195556349224515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,51200,0.04145155681504144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,256,0.008038221961922115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,16384,0.018953777021831937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,16384,0.016978666186332703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,2048,0.010724444356229572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,16384,0.018787556224399142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,12288,0.015327110886573792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,16384,0.031139555904600356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,128,0.0032151111712058387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,51200,0.04526933365397983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,12288,0.01622488929165734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,10240,0.013789333403110504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,12288,0.02572800053490533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,10240,0.013832888669437833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,3072,0.007860444486141205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,8192,0.011958222422334882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,10240,0.022791999909612868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,8192,0.011754666765530905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,8192,0.012977778083748288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,8192,0.020273778173658583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,51200,0.07848711146248712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,12288,0.015498666299713982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,6144,0.010251555177900527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,7168,0.01863822175396813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,10240,0.01500266624821557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,65536,0.056483553515540234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,5120,0.009699555734793345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,6144,0.009684444301658207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,6144,0.017637333936161466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,7168,0.011776888536082374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,5120,0.009008888569143083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,5120,0.008832000195980072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,5120,0.015492444237073263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,7168,0.010384889112578498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,4096,0.007636444436179266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,4096,0.013678222066826291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,3584,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,3584,0.012753778033786349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,3072,0.008067555725574493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,4096,0.009177777502271865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,64,0.0053680001033677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,7168,0.0120666664507654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,4096,0.00832177781396442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,2560,0.007081777685218387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,3072,0.007706666986147563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,6144,0.010839111275143094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,2560,0.006679111056857639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,2048,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,2048,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,3584,0.008447999755541483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,2048,0.006735999965005451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,2048,0.0107022225856781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,3072,0.012470222181744046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,2560,0.011432888607184092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,1536,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,1024,0.004567111117972268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,1536,0.006675555474228329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,1024,0.005632888939645555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,1024,0.009368889033794403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,1536,0.00979733301533593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,768,0.00425155543618732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,1536,0.005778666585683823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,768,0.005961777849329843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,3584,0.008006222546100616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,768,0.009045333498054082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,768,0.005516444643338521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,512,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,512,0.005326222214433882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,512,0.008723555339707268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,3072,0.006994666324721442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,128,0.0029146667155954572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,256,0.0058311112225055695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,1024,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,256,0.008219555848174626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,64,0.0025519999778932994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,128,0.004903111192915174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,32,0.0025333333760499954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,2560,0.00741244438621733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,65536,0.05000177688068814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,64,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,256,0.0034328889515664843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,32,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,51200,0.038458665211995445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,65536,0.050435556305779346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,256,0.0052871112194326185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,51200,0.04007110993067423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,65536,0.09814577632480198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,16384,0.016151999433835346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,51200,0.0786657796965705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,128,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,12288,0.013232889274756113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,16384,0.018560889694425795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,16384,0.01661777827474806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,16384,0.030954665607876245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,10240,0.011468444433477191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,12288,0.014208889669842191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,12288,0.015153777268197803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,12288,0.02571644385655721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,65536,0.055677334467569985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,10240,0.01330044451687071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,51200,0.04526844289567736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,128,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,10240,0.023153778579499986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,7168,0.010832889212502373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,512,0.0038808890514903595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,8192,0.01997866729895274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,8192,0.011748444702890186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,6144,0.009359999663299983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,7168,0.010233778092596266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,7168,0.018239110708236694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,7168,0.010233778092596266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,6144,0.016580444243219163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,8192,0.010936000280910067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,6144,0.00904266701804267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,4096,0.0101724440852801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,5120,0.009015111459626092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,5120,0.008384000096056197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,5120,0.014818666709793938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,8192,0.010775110787815519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,4096,0.013274666335847644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,4096,0.007443555527263218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,3584,0.007290666302045186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,6144,0.009586667021115622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,3584,0.01277599980433782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,10240,0.012488000094890594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,3072,0.006861333217885759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,3072,0.007709333466158972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,2560,0.005614222337802251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,3072,0.012091555529170565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,3584,0.008994667066468133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,2048,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,2560,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,4096,0.008448888858159384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,2560,0.011076444553004371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,1536,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,2048,0.0069955554273393415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,5120,0.008743111458089616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,2048,0.010694222317801582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,1024,0.004899555610285865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,1536,0.010081777969996134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,2560,0.006607999818192587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,3072,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,1024,0.009367999931176504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,768,0.004200889004601372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,1536,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,768,0.005427555491526921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,768,0.00869066682126787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,512,0.0036026665733920205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,3584,0.008007110820876228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,1536,0.006640000061856375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,512,0.008340444829728868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,256,0.0032506665835777917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,1024,0.005628444254398346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,256,0.005119110975000593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,256,0.005640888793600931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,2048,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,256,0.008345777789751688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,1024,0.006321777900060018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,64,0.0028444443725877335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,128,0.005650666852792104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,512,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,32,0.0029093333416514923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,65536,0.036095112562179565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,64,0.005236444373925527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,768,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,32,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,51200,0.0284551117155287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,65536,0.054924445019827954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,512,0.005281777845488654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,65536,0.09805955489476521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,16384,0.013230221966902414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,51200,0.03874311182234023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,51200,0.044396443499459155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,128,0.004893333133724001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,12288,0.011919111013412476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,128,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,16384,0.016325333052211337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,16384,0.030976000759336684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,10240,0.010102222363154093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,12288,0.014318222800890604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,12288,0.025825776987605627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,65536,0.04829244481192695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,10240,0.01276177747382058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,10240,0.012133333418104382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,8192,0.010431110858917236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,8192,0.010038221875826517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,51200,0.07826933595869276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,7168,0.009678222239017487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,8192,0.019528888993793063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,12288,0.013432888521088494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,7168,0.009404444032245213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,6144,0.00888355573018392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,7168,0.01753777762254079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,128,0.0032675556010670135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,6144,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,6144,0.008685333033402761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,6144,0.016399999459584553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,10240,0.022619555393854778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,4096,0.007762666377756331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,5120,0.008037333687146505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,8192,0.010820444259378644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,5120,0.014860444598727755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,3584,0.007342221836249034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,4096,0.007535999847782983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,4096,0.013122667041089801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,3584,0.007318221860461765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,3072,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,3584,0.012764444781674279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,3072,0.006953777538405524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,5120,0.00886044485701455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,3072,0.012087110843923358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,2560,0.005617777920431561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,3072,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,4096,0.008344000412358178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,2560,0.0073315559162033936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,2560,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,7168,0.010391111175219217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,2560,0.011516444385051727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,1536,0.005316444569163852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,3584,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,2048,0.0061039999127388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,2048,0.01070755554570092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,1024,0.004266666869322459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,1536,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,1536,0.00981866651111179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,16384,0.017287110288937885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,768,0.003896000070704354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,1024,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,1024,0.009400889277458191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,2048,0.00598933340774642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,512,0.003585777762863371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,768,0.006000888844331105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,768,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,2048,0.007000000112586551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,256,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,512,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,512,0.008657777474986183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,1536,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,128,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,256,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,256,0.00565244464410676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,256,0.008032889001899296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,1024,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,64,0.002893333426780171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,128,0.00565422202150027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,128,0.004953777624501122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,128,0.008201777935028076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,65536,0.025424000289705064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,64,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,5120,0.007970666719807519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,512,0.005315555466545953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,51200,0.021659556362364028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,65536,0.05453244513935513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,65536,0.09755911429723103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,51200,0.04383288820584615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,16384,0.014919999572965833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,51200,0.037809779246648155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,768,0.005294222384691238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,51200,0.07826489210128784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,16384,0.016197333733240765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,32,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,65536,0.046605332030190356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,12288,0.012838222086429596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,12288,0.013163555827405719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,32,0.0028488888508743713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,10240,0.011243555280897351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,10240,0.012107555237081317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,8192,0.010703999963071613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,16384,0.015152000718646579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,10240,0.02199111051029629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,7168,0.009700444837411245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,16384,0.03125155634350247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,8192,0.00980355590581894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,8192,0.018722666634453666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,10240,0.01143200033240848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,6144,0.009008888569143083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,7168,0.009349333743254343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,7168,0.018055111169815063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,6144,0.009702222214804756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,6144,0.00868888861603207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,6144,0.015899555550681222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,5120,0.008365333080291748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,4096,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,5120,0.014319999350441827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,8192,0.01072533345884747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,4096,0.008008889026112026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,7168,0.010066666536860997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,4096,0.013110222087966071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,5120,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,3584,0.007375111182530721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,3072,0.0063386667105886675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,3584,0.012447111308574677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,12288,0.01255022237698237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,3072,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,2560,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,12288,0.025360888904995386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,3584,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,2560,0.006515555497672822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,2048,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,2560,0.011408888631396823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,3584,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,5120,0.00904444439543618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,1536,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,2048,0.006362666686375936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,2048,0.010718222293588849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,3072,0.011746666497654386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,1024,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,4096,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,1536,0.009745777481132084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,2560,0.007070221834712558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,1024,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,1024,0.009350222018029954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,768,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,2048,0.00702311098575592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,3072,0.006993778049945831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,512,0.003579555700222651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,1536,0.006673777682913675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,1536,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,256,0.0032177778581778207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,512,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,512,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,512,0.00871733327706655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,768,0.0042239998777707415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,256,0.005626666463083691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,256,0.004962666581074397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,256,0.008367111285527548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,64,0.002573333266708586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,128,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,32,0.0028293333533737394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,768,0.009027555584907532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,768,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,768,65536,0.021721776988771226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,64,0.004953777624501122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,32,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,768,51200,0.017872888180944655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,65536,0.04598399996757507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,128,0.0029297777348094513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,51200,0.03707022137112088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,51200,0.04386311107211643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,128,0.005633777628342311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,51200,0.07782933447096083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,128,0.008341333104504479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,16384,0.015853333804342482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,768,12288,0.011207111179828644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,65536,0.054229332341088184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,1024,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,12288,0.012096000214417776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,768,10240,0.010011555420027839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,12288,0.02474133339193132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,12288,0.01331555595000585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,10240,0.011066666907734342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,768,8192,0.009233777721722921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,10240,0.01219733307758967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,10240,0.02200444373819563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,16384,0.030478222502602473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,768,7168,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,8192,0.010785777535703448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,8192,0.010050666828950247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,65536,0.09752000040478176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,7168,0.010386666489972008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,7168,0.009358222285906473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,7168,0.017799999978807237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,6144,0.008790222307046255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,768,5120,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,6144,0.015857777661747403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,16384,0.014478221535682678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,768,4096,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,5120,0.00817599975400501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,4096,0.007613333563009898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,768,6144,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,768,3584,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,768,16384,0.010390222072601318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,8192,0.019343111250135634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,3584,0.008003555238246918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,768,3072,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,5120,0.008671110702885522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,5120,0.014817777607176038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,3072,0.006990222467316522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,768,2560,0.0063324446479479475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,3072,0.011764444410800934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,4096,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,4096,0.013591999808947245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,768,2048,0.005081777771313985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,2560,0.007300444775157505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,2560,0.011041777829329172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,3584,0.012454222473833295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,768,1536,0.004912000149488449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,2048,0.010791111323568555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,3072,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,1536,0.0059262220230367445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,1536,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,1536,0.010047111246320937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,2560,0.006589333216349284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,1024,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,768,768,0.003879111260175705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,1024,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,2048,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,2048,0.006334222025341458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,768,512,0.003301333429084884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,768,0.005984000033802456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,768,0.005320888840489917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,512,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,768,1024,0.004261333495378494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,512,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,512,0.005324444423119227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,6144,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,768,128,0.0028435554769304064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,256,0.004935111022657818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,1024,0.009408888717492422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,128,0.005635555419656966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,768,64,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,128,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,768,32,0.002595555658141772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,768,256,0.0031893334041039148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,512,65536,0.017095999585257638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,32,0.004959110998445087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,3584,0.007293333609898885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,256,0.005701333284378052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,512,51200,0.015012444721327888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,65536,0.04549422197871738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,65536,0.09728266795476277
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,128,0.004969777746333016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,512,16384,0.010830221904648675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,51200,0.03667466508017646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,51200,0.07794755697250366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,768,0.009100444614887238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,16384,0.014632888966136508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,512,12288,0.008838222258620793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,16384,0.031102223528756037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,12288,0.012126222252845764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,65536,0.04792799883418613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,512,10240,0.007991111112965478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,12288,0.013578666581047906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,12288,0.024634665913052026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,51200,0.03879911038610671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,512,8192,0.00813688917292489
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,10240,0.02201955517133077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,64,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,512,7168,0.007802666889296637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,16384,0.01583377851380242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,8192,0.019283556275897555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,256,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,7168,0.01754844519827101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,7168,0.010046222143703038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,512,6144,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,10240,0.01106844428512785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,10240,0.012139555480745105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,512,5120,0.007035555938879649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,6144,0.008672888908121321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,8192,0.010767999622556897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,8192,0.009696889254781935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,6144,0.016390222642156813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,5120,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,5120,0.014480888843536377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,7168,0.009359999663299983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,512,3584,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,4096,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,4096,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,4096,0.013425778183672162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,512,3072,0.006637333167923822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,3584,0.008031110796663497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,3584,0.007088888850477006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,6144,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,512,2560,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,3072,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,512,4096,0.007279111279381647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,3072,0.011759111450778114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,512,2048,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,2560,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,2560,0.011593778100278644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,5120,0.008405333591832055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,2048,0.006960888703664144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,2048,0.006367110957702001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,2048,0.010392000277837118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,512,1024,0.0038844446341196695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,1536,0.010022222167915767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,1536,0.006689777804745569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,1536,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,512,768,0.0036319999231232535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,1024,0.005352889084153705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,1024,0.0063786668082078295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,2560,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,1024,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,512,512,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,512,1536,0.004591111093759537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,768,0.008965333302815756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,768,0.005320000151793162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,512,256,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,512,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,512,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,512,0.008678221868144141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,256,0.005706666658322017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,256,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,256,0.007668444679843054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,512,64,0.0028479999552170434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,3072,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,512,32,0.002534222271707323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,128,0.008003555238246918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,256,65536,0.01566488875283135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,64,0.004564444637960858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,768,0.0064088888466358185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,65536,0.04265066650178698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,65536,0.0466231107711792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,512,128,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,65536,0.0964311096403334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,3584,0.012430222498046027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,51200,0.03462400039037069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,128,0.00564444437623024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,51200,0.0772835546069675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,16384,0.015841777125994366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,256,12288,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,32,0.004952888935804367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,16384,0.014579556054539151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,256,51200,0.013219555219014486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,12288,0.012077333198653327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,256,10240,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,12288,0.02476088868247138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,128,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,10240,0.011076444553004371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,10240,0.02164799968401591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,8192,0.009915555516878763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,8192,0.018597332967652213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,8192,0.01072088877360026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,16384,0.030470222234725952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,12288,0.013457777599493662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,7168,0.009360888765917884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,256,16384,0.008348444269763099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,7168,0.010106667048401302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,7168,0.017826666434605915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,6144,0.00868888861603207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,256,5120,0.00630044440428416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,6144,0.00961688905954361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,256,7168,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,6144,0.016320889194806416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,256,4096,0.005771555420425203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,5120,0.008652444514963362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,51200,0.037412444750467934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,256,8192,0.00700444479783376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,256,3584,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,5120,0.015075556106037565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,4096,0.013449778159459433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,4096,0.007385777930418651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,3584,0.00812888890504837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,3584,0.0074275558193524676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,5120,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,10240,0.012074666718641916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,3072,0.011778666741318174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,3072,0.007678222325113084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,2560,0.006330666856633292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,3584,0.012763555679056378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,256,3072,0.005971555494599872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,256,2048,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,2560,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,2560,0.01143200033240848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,256,6144,0.006657777975002925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,256,1536,0.004574222283230888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,2048,0.006952889263629913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,2048,0.010759111079904767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,1536,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,1536,0.006002666635645761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,1536,0.010173333187898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,1024,0.005618666609128316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,1024,0.00960711141427358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,3072,0.007117333511511485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,2048,0.006269333263238271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,4096,0.008098666866620382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,256,512,0.003195555466744635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,768,0.006340444501903322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,256,1024,0.0038773334688610504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,768,0.008806222014957005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,512,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,256,256,0.003031999907559819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,256,768,0.003547555456558863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,256,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,256,2560,0.006292444550328785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,768,0.0053173332578606075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,128,0.0046453335218959385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,512,0.004951111144489712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,128,0.008386666576067606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,128,0.005622222191757626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,256,32,0.002542222125662698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,1024,0.006334222025341458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,64,0.004949333353175057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,128,65536,0.014204444156752693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,32,0.004589333302444882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,256,0.008384000096056197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,128,51200,0.012874667015340594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,65536,0.0421688887808058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,65536,0.09626933601167466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,51200,0.034295999341540866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,256,64,0.0025902222841978073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,128,16384,0.008066666622956594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,51200,0.07662044631110297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,512,0.00848444468445248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,16384,0.01588622232278188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,128,12288,0.00739644467830658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,16384,0.030299554268519085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,12288,0.012159110771285163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,128,10240,0.007002666592597961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,12288,0.013553777502642738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,12288,0.024480889240900677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,51200,0.03734044565094842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,10240,0.011095110740926532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,10240,0.012178666889667511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,16384,0.014182221558358936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,10240,0.021773333350817364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,8192,0.009738666315873465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,128,7168,0.006738666858938005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,256,128,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,65536,0.04594755503866407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,128,6144,0.006066666709052191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,8192,0.01901155544651879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,128,8192,0.006804444309737947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,256,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,128,5120,0.00592533333433999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,8192,0.01070844464831882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,5120,0.00832088871134652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,7168,0.009402666654851701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,128,4096,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,5120,0.014144000079896716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,6144,0.008623999853928884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,4096,0.007367110914654202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,6144,0.009391110804345872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,128,3584,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,6144,0.016195555528004963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,3584,0.008015111088752747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,3584,0.006981333096822103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,7168,0.010087110930018956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,5120,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,3072,0.006683555742104848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,128,2560,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,3072,0.007667555577225155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,4096,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,4096,0.012831110921170978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,2560,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,2560,0.00702044450574451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,128,3072,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,3584,0.012648888760142855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,128,1536,0.004254222330119875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,2048,0.006960888703664144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,2048,0.010394666757848527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,3072,0.012016889121797351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,1536,0.005964444329341252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,128,1024,0.003567111161020067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,128,2048,0.005614222337802251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,2560,0.011048888994587792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,1024,0.005607999861240387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,128,768,0.003556444413132138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,2048,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,1024,0.009003555609120263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,768,0.005964444329341252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,7168,0.018049778209792245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,128,512,0.0031928889867332247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,1536,0.006647999915811751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,512,0.005292444593376583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,128,256,0.0028942223224374983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,512,0.008371555142932469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,1024,0.006263111200597551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,256,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,256,0.005144888742102517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,128,128,0.0028391112056043413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,768,0.0053680001033677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,768,0.009022222624884712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,128,0.004991111242108875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,128,64,0.002838222309947014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,128,0.006839111033413146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,512,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,64,65536,0.010763555765151978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,64,51200,0.00996888925631841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,32,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,64,16384,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,65536,0.0421324438518948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,64,12288,0.006454222318198945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,16384,0.01443822185198466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,64,10240,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,1536,0.009535111486911774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,12288,0.012109333442317115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,64,8192,0.00600000015563435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,128,32,0.0025928889711697898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,64,7168,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,64,6144,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,10240,0.011051555474599203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,64,5120,0.005640888793600931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,8192,0.009762666291660732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,7168,0.0090515555606948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,6144,0.00833155545923445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,64,3584,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,256,0.007679111427730984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,64,3072,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,128,0.005630222045713001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,3584,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,3072,0.006599111275540457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,64,2048,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,2560,0.006265777680608962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,64,1536,0.0042408886882993905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,64,4096,0.005616888817813661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,64,1024,0.00360622215602133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,5120,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,1536,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,4096,0.007359999749395583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,1024,0.005647999958859549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,64,2560,0.006049777898523543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,768,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,512,0.005246222019195557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,64,128,0.002598222138153182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,64,0.004974222017659081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,64,64,0.0025439999169773525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,2048,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,64,32,0.002526222210791376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,51200,0.03380977776315477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,64,768,0.003526222374704149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,64,0.004987555659479565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,64,512,0.0031964443624019623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,32,51200,0.008384889198674096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,32,0.004573333180612988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,32,16384,0.006350222147173352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,32,12288,0.005976000179847081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,65536,0.041830221811930336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,32,10240,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,16384,0.014157333307796054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,32,8192,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,12288,0.01219555570019616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,32,7168,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,128,0.004938666605287128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,32,6144,0.0063368889192740125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,32,65536,0.009506666825877296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,32,5120,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,7168,0.009353777600659264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,32,4096,0.005631111148330901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,6144,0.008654221892356873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,32,3584,0.005630222045713001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,5120,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,51200,0.03382044368320041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,32,3072,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,32,2560,0.006220444622966979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,256,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,3584,0.007027555671003129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,10240,0.011040888726711273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,32,1536,0.004231111043029361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,2560,0.00646577775478363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,8192,0.009672889278994666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,32,1024,0.0038737778862317405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,1536,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,32,768,0.0035537779331207275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,1024,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,32,512,0.0028773333049482773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,4096,0.007297777467303806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,768,0.0053031109273433685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,32,2048,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,512,0.004936000125275718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,256,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,2048,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,128,0.004583111239804162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,32,32,0.0022888889329300984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,64,256,0.0028799999919202593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,64,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,32,0.004581333448489507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,32,256,0.002926222152180142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,3072,0.006673777682913675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,16384,0.34274665514628094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,32,64,0.0028702221396896574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,51200,1.0218097898695204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,16384,0.24470755789015028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,12288,0.26202665434943306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,16384,0.13404533598158094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,51200,0.43242488967047793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,8,32,128,0.0026622222115596137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,10240,0.15452622042761907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,12288,0.1523617770936754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,8192,0.15498044755723742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,51200,0.4863315688239203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,10240,0.0859137773513794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,7168,0.13292000028822157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,8192,0.07128266493479411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,8192,0.14021955596076116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,6144,0.11491289403703477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,5120,0.09667555491129558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,6144,0.05510310994254219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,6144,0.086535996860928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,4096,0.08072533210118611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,5120,0.07426755295859443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,12288,0.11558132701449925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,3584,0.06881688700781928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,4096,0.03939377930429246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,3072,0.0605857769648234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,3584,0.03497155507405599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,3584,0.05446755554940966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,2560,0.050986667474110924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,3072,0.030937777625189886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,7168,0.09803111023373073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,2048,0.04204710986879137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,2560,0.02717511190308465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,2560,0.039228445953793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,5120,0.04709688822428385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,2048,0.023010666171709698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,2048,0.035882665051354304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,4096,0.06519289149178399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,1536,0.018914666440751817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,1024,0.01792888840039571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,1024,0.020008888509538438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,768,0.018782221608691745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,3072,0.05227466755443149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,512,0.013430222041077085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,7168,0.06269511249330309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,10240,0.2061386638217502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,768,0.016807110773192514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,1536,0.032484445306989886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,512,0.01310222182008955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,128,0.006265777680608962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,1024,0.023725334140989516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,256,0.01036177741156684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,1536,0.027451554934183758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,128,0.010423999693658618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,32,0.00592533333433999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,64,0.00902844468752543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,768,0.01519377695189582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,256,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,512,0.013183111117945777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,256,0.010712888505723743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,64,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,65536,0.9696746402316623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,51200,0.7509973314073352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,16384,0.2516728772057427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,32,0.009350222018029954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,51200,0.38933600319756406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,12288,0.19297422303093803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,65536,0.42334310213724774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,16384,0.14662044578128391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,12288,0.14307199584113225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,51200,0.31685598691304523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,10240,0.1623964442147149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,12288,0.08458489179611206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,8192,0.13956711027357313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,10240,0.10268444485134548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,8192,0.0588959985309177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,8192,0.09279555744594997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,6144,0.10643111334906684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,7168,0.0772248903910319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,6144,0.044770667950312294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,6144,0.07136088609695435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,5120,0.09050666623645359
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,16384,0.12771111064487034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,10240,0.07228800323274401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,7168,0.1242026620441013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,5120,0.060552000999450684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,7168,0.051966223451826304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,4096,0.03291822142071194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,3584,0.06372000111473931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,4096,0.05396177702479892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,3072,0.055993775526682533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,128,0.008027555214034187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,3584,0.045237332582473755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,2560,0.04684888985421923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,3072,0.025780444343884785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,3072,0.038077334562937416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,2048,0.038005332152048744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,2560,0.022711111439598933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,5120,0.03907288776503669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,2560,0.032447109619776406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,1536,0.029469334416919287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,2048,0.01941422290272183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,2048,0.030184000730514526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,1536,0.01680888897842831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,1536,0.022876444790098403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,768,0.016346666547987197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,65536,0.5138559871249729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,1024,0.01698755555682712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,768,0.013611555927329592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,768,0.01519555515713162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,256,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,512,0.011527111132939657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,4096,0.0743466681904263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,512,0.011376889215575324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,128,0.0039022221333450745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,256,0.009696000152164036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,64,0.003212444484233856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,128,0.009686222506894005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,32,0.0035413333939181436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,128,0.007303111255168915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,64,0.008278222547637092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,3584,0.03499555587768555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,32,0.008386666576067606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,65536,0.32199109925164116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,51200,0.2502933343251546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,65536,0.17896533012390137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,65536,0.15551911460028753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,16384,0.08709777726067437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,51200,0.13055645094977483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,256,0.00850666645500395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,12288,0.06731288962894015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,1024,0.021524444222450256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,51200,0.12905155287848577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,10240,0.05498844385147095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,12288,0.04968355430497063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,512,0.01147911118136512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,10240,0.0469688872496287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,8192,0.024937777055634394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,7168,0.03716889023780823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,8192,0.036581334140565656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,16384,0.042071110672420926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,16384,0.06521599822574191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,6144,0.03236088818973965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,12288,0.03865955604447259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,5120,0.028011557128694322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,6144,0.01997066703107622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,6144,0.027775110469924078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,4096,0.024096000525686476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,5120,0.01776977711253696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,5120,0.024884444144037034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,3584,0.020658666888872784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,4096,0.02125511070092519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,3584,0.017654221918847825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,7168,0.021946666969193354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,1024,0.01575555569595761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,10240,0.028466665082507666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,2560,0.015499555402331881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,3072,0.0162453336848153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,2560,0.012295111185974546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,4096,0.015826667348543804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,2048,0.013279111021094851
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,2048,0.011113777756690979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,3584,0.018606222338146634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,3072,0.018565333551830716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,8192,0.043544888496398926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,1536,0.009365333451165093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,1024,0.008208889100286696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,1536,0.011667555405033959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,768,0.0069422221018208405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,1024,0.009319999979601966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,1024,0.009411555197503831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,2560,0.014797333214018079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,512,0.005167110926575131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,768,0.008024888734022776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,512,0.007359999749395583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,512,0.0069573331210348345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,256,0.006959111326270633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,256,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,128,0.003185777821474605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,3072,0.01646577815214793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,128,0.006949333681000604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,64,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,128,0.005330666485759947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,32,0.002895111011134254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,768,0.008360000120268928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,64,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,65536,0.2456080118815104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,1536,0.010805333654085795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,7168,0.03189244535234239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,65536,0.14539021915859646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,65536,0.1327377822664049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,16384,0.06676977872848511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,51200,0.11184444692399766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,51200,0.10672177871068318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,12288,0.05168977710935804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,16384,0.03775377737151252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,16384,0.05324177940686544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,2048,0.014001778430408902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,10240,0.0445128877957662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,12288,0.032102220588260226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,12288,0.04296977652443779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,8192,0.03678755627738105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,10240,0.025958221819665697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,10240,0.037019554111692644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,8192,0.02236977716286977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,8192,0.032280001375410296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,6144,0.028655999236636694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,7168,0.02029333346419864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,7168,0.027385777897304956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,5120,0.02539288832081689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,256,0.0035377778112888336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,6144,0.02404266595840454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,4096,0.02237244447072347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,5120,0.016155555844306946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,5120,0.021174222230911255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,3584,0.019398222366968792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,4096,0.014246222045686511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,7168,0.03240177697605557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,3584,0.015488000379668342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,3584,0.017313778400421143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,2560,0.015489778584904142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,3072,0.014745778507656522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,3072,0.014908444550302295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,2048,0.013545778062608508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,2560,0.012442666623327466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,6144,0.018388445178667705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,2560,0.013439999686347114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,32,0.005920889063013925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,2048,0.00998577806684706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,1024,0.0069751110341813825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,2048,0.013042666845851474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,1536,0.008671999805503422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,51200,0.18997332784864637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,768,0.005881777654091517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,1024,0.008370666868156856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,4096,0.019719110594855416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,512,0.003605333467324575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,768,0.007806222471925948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,256,0.0032035555276605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,512,0.007348444726732042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,512,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,128,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,256,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,1536,0.008934222161769867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,64,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,128,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,128,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,1536,0.01072533345884747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,64,0.0058782220714622075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,3072,0.017294221454196505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,1024,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,32,0.005873777800136142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,768,0.00795377790927887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,65536,0.14294400480058458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,51200,0.1072382198439704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,16384,0.059060447745853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,51200,0.10164178080028957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,12288,0.04385688900947571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,32,0.0028453332682450614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,16384,0.04725333385997348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,10240,0.03797511259714762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,12288,0.028339554866154987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,12288,0.03733155462476943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,8192,0.032010667853885226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,51200,0.1618444389767117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,10240,0.030976888206270006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,65536,0.1280862225426568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,7168,0.028768890433841284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,8192,0.020689777202076383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,8192,0.026927111877335444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,6144,0.0258577780591117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,7168,0.018992889258596633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,5120,0.02218311031659444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,16384,0.036144001616372004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,6144,0.01714666684468587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,4096,0.01995022263791826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,5120,0.015970667203267414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,5120,0.018216888109842937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,65536,0.20677867200639513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,3584,0.017484444710943434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,10240,0.024717332588301763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,4096,0.016410667035314772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,3072,0.01572711103492313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,3584,0.014092443717850579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,3584,0.014827556080288358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,3072,0.013069333301650153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,3072,0.012990222209029727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,7168,0.022878222995334204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,2048,0.010303110712104374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,2560,0.010953777366214328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,6144,0.020623111062579684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,1536,0.009016888837019602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,2048,0.009027555584907532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,4096,0.013990221752060784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,1024,0.006262222097979651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,1536,0.009487999810112847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,256,0.006945778098371293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,768,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,1024,0.008008889026112026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,2560,0.014470222923490735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,512,0.003528888854715559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,768,0.007517333659860823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,768,0.0070017774899800616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,256,0.0032142222755485107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,2560,0.012130666938092975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,512,0.006299555715587404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,128,0.0028488888508743713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,256,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,2048,0.011170667078759937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,256,0.00591111100382275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,1536,0.008126222425036961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,128,0.006268444574541516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,128,0.005492444253630108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,64,0.005621333503060871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,65536,0.17337511645423043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,32,0.005581333405441708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,51200,0.1283759938346015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,512,0.007061333292060428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,65536,0.094595558113522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,16384,0.04515555500984192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,51200,0.068686220380995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,51200,0.056814220216539174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,12288,0.03311199943224589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,64,0.0026791110220882627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,16384,0.022481777601771887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,16384,0.03472977876663208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,12288,0.019080888893869188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,8192,0.02366488840844896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,10240,0.01643200053109063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,10240,0.023970666858885024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,1024,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,7168,0.020503110355801053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,8192,0.022652443912294176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,6144,0.018437332577175565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,7168,0.013814222481515674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,7168,0.017646221650971305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,5120,0.015846222639083862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,6144,0.012855110896958245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,6144,0.016569778323173523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,4096,0.013281777501106262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,5120,0.011617778076065911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,5120,0.014191110928853353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,12288,0.027649778458807204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,4096,0.01070666644308302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,4096,0.013777777552604675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,3072,0.010965333216720156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,3584,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,3584,0.011736888852384357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,2560,0.009464888936943477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,65536,0.07027910815344916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,3072,0.011747555600272285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,2048,0.008347555167145198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,2560,0.010013333625263637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,10240,0.027443556321991816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,1536,0.007098666495747036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,2048,0.007660444411966536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,2048,0.009352000223265754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,1024,0.005584888988071018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,1536,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,1536,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,768,0.003932444585694207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,1024,0.006718222051858902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,3072,0.00924622184700436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,512,0.0035662220584021676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,768,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,768,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,256,0.0032035555276605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,512,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,32,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,128,0.0028826666788922418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,256,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,3584,0.011949333051840464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,256,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,128,0.006004444426960415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,32,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,128,0.004990222139490975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,1024,0.006945778098371293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,64,0.005245333330498801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,65536,0.14835200044843885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,32,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,51200,0.11741688516404893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,2560,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,512,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,51200,0.05375644564628601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,51200,0.06247111161549886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,8192,0.014721777704026965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,16384,0.021756443712446425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,12288,0.033131556378470525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,12288,0.01824977828396691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,12288,0.026713778575261433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,10240,0.028578665521409776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,65536,0.08316266536712646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,8192,0.024295111497243244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,10240,0.015961777832772996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,10240,0.022036444809701707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,16384,0.041478223270840116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,7168,0.018886221779717337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,8192,0.014207111464606391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,8192,0.019938665959570143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,6144,0.01679022279050615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,7168,0.013146667016877068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,16384,0.03314044409328037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,5120,0.014489778214030795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,4096,0.012703999876976013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,5120,0.01108444482088089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,65536,0.06606133116616143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,4096,0.009869332942697737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,4096,0.012431110772821637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,64,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,3584,0.01145333300034205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,3072,0.010036444498433007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,3072,0.008341333104504479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,6144,0.012401777837011548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,2560,0.009026666482289633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,6144,0.015541333291265698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,2048,0.007965332931942409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,2560,0.007970666719807519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,5120,0.014145778285132514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,3584,0.011063111325105032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,2048,0.008900444540712569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,3584,0.008742222355471717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,1024,0.004297777596447202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,1536,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,7168,0.016708445217874315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,768,0.003943110919660992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,1024,0.006326222171386083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,2560,0.009851555857393477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,1536,0.006601777755551868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,512,0.005658666706747479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,512,0.006233777850866318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,1024,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,256,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,3072,0.010449777874681683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,128,0.002867555452717675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,256,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,256,0.005230222311284807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,768,0.006320000108745363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,64,0.0028844444702068963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,512,0.0032426667296224167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,32,0.0028497777465316984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,128,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,65536,0.12183466222551133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,64,0.00526577772365676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,1536,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,51200,0.0971226692199707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,65536,0.07632711198594835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,65536,0.06327466832266913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,16384,0.035573333501815796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,51200,0.05634311172697279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,16384,0.021405332618289526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,12288,0.02863822380701701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,16384,0.030510223574108545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,10240,0.024671110841963027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,128,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,2048,0.007598222129874759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,12288,0.024318221542570326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,10240,0.015417777829700045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,10240,0.020581333173645865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,32,0.0052968888647026485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,8192,0.013554666605260638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,8192,0.018844443890783522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,6144,0.01703377730316586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,7168,0.012811555630630918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,7168,0.016306666864289176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,5120,0.015167999598715039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,6144,0.01514488955338796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,768,0.006047111004590988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,5120,0.010670222342014313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,4096,0.012025777664449481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,4096,0.00906755526860555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,7168,0.01921066641807556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,3584,0.010282666318946415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,4096,0.0124195557501581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,3072,0.009279111193286048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,3584,0.008584889272848764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,3584,0.01108000013563368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,2560,0.00830844458606508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,3072,0.009858667022652095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,12288,0.017256889078352187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,2048,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,51200,0.05128088924619886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,6144,0.011746666497654386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,2048,0.007294221884674496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,2560,0.008007110820876228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,1536,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,2560,0.008609777523411645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,1024,0.004222222086456087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,1536,0.006993778049945831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,1536,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,768,0.003559111307064692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,1024,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,768,0.006304889089531369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,768,0.006222222414281633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,2048,0.008113777471913232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,512,0.005579555614127054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,256,0.0031991110493739447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,512,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,256,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,128,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,3072,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,128,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,128,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,32,0.002569777684079276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,5120,0.013370667066839008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,64,0.0052782222628593445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,65536,0.10677422417534722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,32,0.0053075556125905775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,256,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,51200,0.08455822202894424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,65536,0.0662720004717509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,16384,0.03173333406448364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,512,0.003206222214632564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,51200,0.049572444624371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,12288,0.02587822245226966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,16384,0.030961778428819444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,8192,0.021344888541433547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,10240,0.02180266711446974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,12288,0.016765332884258695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,12288,0.022385777698622808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,8192,0.01961777773168352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,10240,0.015167999598715039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,10240,0.022337777747048274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,7168,0.017300445172521803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,8192,0.017800889081425138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,64,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,6144,0.01567555632856157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,16384,0.020267556111017864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,6144,0.011395555403497485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,5120,0.013809777796268463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,6144,0.013842666314707862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,1024,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,5120,0.009961778091059791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,65536,0.061258667045169406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,5120,0.013101333545313941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,8192,0.013395555317401886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,3584,0.00944622192117903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,4096,0.008723555339707268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,3072,0.00870222184393141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,51200,0.05080355538262261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,7168,0.012494222157531314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,3584,0.010483555495738983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,2560,0.007673777639865875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,3072,0.008026666939258575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,2048,0.0069351109365622205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,3072,0.009191110730171204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,2560,0.007339555356237624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,1536,0.005504000104135937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,2048,0.00721155520942476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,2048,0.007648000286685095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,1024,0.0038888889054457345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,1536,0.006984000404675801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,4096,0.01184977756606208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,1024,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,768,0.0035333335399627686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,7168,0.015825778245925903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,512,0.0032586666444937387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,768,0.006308444258239534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,768,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,256,0.002899555489420891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,2560,0.0081404447555542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,512,0.005624889085690181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,128,0.002569777684079276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,256,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,256,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,64,0.0028737777223189673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,128,0.004903111192915174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,1024,0.006365333166387346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,32,0.0028702221396896574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,65536,0.08995288610458374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,32,0.005315555466545953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,512,0.006014222072230444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,65536,0.05906399753358629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,4096,0.011108444796668159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,1536,0.007009777757856581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,16384,0.026194666822751362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,128,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,51200,0.04653155472543505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,3584,0.008260444634490544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,16384,0.0196079992585712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,16384,0.02588622272014618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,64,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,10240,0.01827555563714769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,12288,0.022316444251272414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,10240,0.014463111758232117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,10240,0.01759022143152025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,65536,0.058873779243893094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,8192,0.012811555630630918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,7168,0.014333332578341166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,12288,0.02068088948726654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,7168,0.011744889120260874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,12288,0.016197333733240765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,6144,0.01315111087428199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,6144,0.010461333725187512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,51200,0.06684888733757867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,6144,0.013109332985348172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,5120,0.009733333355850643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,51200,0.04758488800790575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,5120,0.011743110915025076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,8192,0.016262221667501662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,3584,0.012060444388124677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,4096,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,4096,0.011052444577217102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,3072,0.010197333163685268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,3584,0.009310222334331935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,3072,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,8192,0.016128000285890367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,4096,0.010815111299355825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,2048,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,2560,0.00738666703303655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,2048,0.00699733363257514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,2048,0.007154666715198093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,1536,0.006326222171386083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,1536,0.006704000135262807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,5120,0.01202400028705597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,1024,0.004950222041871813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,2560,0.008736888567606607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,768,0.0042151109211974675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,1024,0.006335999816656113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,1024,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,512,0.0038888889054457345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,7168,0.014205333259370593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,512,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,256,0.0028702221396896574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,512,0.00599911105301645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,1536,0.006643555644485686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,256,0.005976888868543837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,256,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,2560,0.007640889121426477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,64,0.002867555452717675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,128,0.004966222163703707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,32,0.0028213332924577924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,3584,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,64,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,65536,0.07877422041363187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,32,0.005244444641802046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,51200,0.06111288732952542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,3072,0.008997333546479544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,128,0.0027253333893087176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,16384,0.024319110645188227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,51200,0.04683644572893778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,128,0.0059164443777667145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,768,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,16384,0.019630221856964957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,16384,0.02534488836924235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,10240,0.016751999656359356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,12288,0.016128000285890367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,12288,0.020285333196322124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,8192,0.014672888649834527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,10240,0.014449778530332776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,10240,0.016833777228991192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,7168,0.013732444081041547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,65536,0.05782666471269396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,8192,0.016161777906947665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,6144,0.012436444560686747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,7168,0.011056889262464313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,768,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,7168,0.013016000390052795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,5120,0.011126221881972419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,6144,0.009929777847396003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,4096,0.010392888552612728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,6144,0.012706666356987424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,5120,0.00907022257645925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,3584,0.011081777513027191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,4096,0.010052444206343757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,8192,0.012424888710180918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,65536,0.05743822124269274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,3072,0.010026666853162978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,3584,0.009054222040706212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,2560,0.008638222184446123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,51200,0.04476266768243578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,12288,0.01898755629857381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,3072,0.007974222302436829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,2048,0.007359999749395583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,2560,0.007378666765160031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,1536,0.005994666781690385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,2048,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,2048,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,1024,0.004662222332424588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,4096,0.008575999902354347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,1536,0.006641777853171031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,1024,0.006304889089531369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,768,0.004212444441186057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,3072,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,768,0.0058204444746176405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,768,0.006266666783226862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,512,0.003896000070704354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,2560,0.007344889144102733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,256,0.002887111157178879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,512,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,1536,0.006675555474228329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,512,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,128,0.0026782221264309357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,256,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,256,0.004949333353175057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,3584,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,1024,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,128,0.004909333255555895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,65536,0.06267644299401177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,64,0.004942222187916438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,32,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,51200,0.05049066742261251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,65536,0.05451644460360209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,5120,0.011343110766675739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,16384,0.020633776982625324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,51200,0.04592355423503452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,51200,0.04279555545912849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,64,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,12288,0.017141333884663053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,128,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,16384,0.019076445036464267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,32,0.0028799999919202593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,16384,0.02402488887310028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,12288,0.015838222371207345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,8192,0.013438222308953604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,12288,0.018181333939234417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,7168,0.012356444365448423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,10240,0.016467556357383728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,8192,0.014862222803963555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,6144,0.011406222151385413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,7168,0.012648888760142855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,5120,0.01071200023094813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,6144,0.011992000043392181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,10240,0.014921777778201632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,5120,0.009769777456919352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,4096,0.009685333404276106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,10240,0.013909333282046847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,4096,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,8192,0.01183999992079205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,4096,0.008616000413894653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,65536,0.057251552740732826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,3584,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,3584,0.008004444340864817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,2560,0.007801777786678738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,3072,0.007359999749395583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,2560,0.007365333537260692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,5120,0.009032888544930352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,3584,0.009348444640636444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,2048,0.006991110742092133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,1536,0.006016888966162999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,2048,0.0069875551594628235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,7168,0.010848888920413123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,1024,0.004944889081848992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,1536,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,3072,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,1536,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,1024,0.0062977779242727495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,1024,0.005665777872006099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,512,0.0038853333228164246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,2560,0.0069546666410234236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,768,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,256,0.003535110917356279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,512,0.005980444451173146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,512,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,128,0.00318755561278926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,256,0.005767111149099138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,256,0.005215999980767568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,64,0.0025475554996066624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,6144,0.009692444569534725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,32,0.002838222309947014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,64,0.005092444519201915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,2048,0.00664533343580034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,65536,0.05684000253677368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,65536,0.05628799729877048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,3072,0.008664888640244802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,51200,0.04478222131729126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,128,0.0059262220230367445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,16384,0.018637332651350234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,51200,0.04541155695915222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,51200,0.04130666785769992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,12288,0.015856888559129503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,16384,0.018955555227067735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,16384,0.021373333202468023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,32,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,10240,0.013833777772055732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,12288,0.015562666787041558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,12288,0.017038222816255357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,8192,0.012143111063374413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,10240,0.01314222233162986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,7168,0.01145333300034205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,10240,0.014852444330851236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,8192,0.013517333401574029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,128,0.004907555464241239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,7168,0.010767111347781287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,768,0.00424177779091729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,6144,0.009692444569534725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,768,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,6144,0.010445333189434476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,5120,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,4096,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,5120,0.008961777720186446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,8192,0.010787555740939246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,3584,0.009020444419648912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,4096,0.008368888662921058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,4096,0.00798044436507755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,7168,0.011957333319716983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,3072,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,3584,0.008044444852405125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,2560,0.007292444507280986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,3072,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,5120,0.009692444569534725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,3072,0.006980444822046492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,2048,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,2560,0.007354666789372762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,1536,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,2048,0.006960888703664144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,65536,0.05264266663127475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,1024,0.004939555707905027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,1536,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,1024,0.006522666662931442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,3584,0.00797333319981893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,768,0.004474666797452503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,768,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,512,0.0038888889054457345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,768,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,2560,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,512,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,2048,0.006251555350091722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,1536,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,256,0.005728000154097875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,256,0.005121777868933148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,64,0.002529777793420686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,128,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,6144,0.010695111420419482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,32,0.0025253333151340485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,65536,0.04929688904020521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,64,0.0052382221652401825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,32,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,51200,0.038410666916105486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,65536,0.05549600058131748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,512,0.0052791109515561005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,128,0.0032151111712058387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,51200,0.039975110027525164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,16384,0.018573333819707234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,1024,0.005624889085690181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,12288,0.013057777451144325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,12288,0.014736889137162102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,10240,0.011724444727102915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,12288,0.015473778049151102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,256,0.003549333247873518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,8192,0.011277333729796939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,10240,0.013797333670987023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,16384,0.015549333559142219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,7168,0.010703999963071613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,8192,0.011035555766688453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,8192,0.01257333325015174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,6144,0.009759110709031424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,128,0.004945777770545748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,7168,0.01038933296998342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,5120,0.009002666506502364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,6144,0.009007111191749573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,65536,0.0499253339237637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,5120,0.008834666675991481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,4096,0.010268444816271463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,51200,0.04486844605869717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,4096,0.00833777752187517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,4096,0.007678222325113084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,16384,0.02067377832200792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,3584,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,6144,0.009675555759006077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,3072,0.007660444411966536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,3072,0.0070257774657673305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,2560,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,5120,0.008345777789751688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,10240,0.012200000385443369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,2048,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,2560,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,1536,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,2048,0.007119999991522894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,2048,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,1536,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,3584,0.008007110820876228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,1536,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,3072,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,768,0.004584889031118817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,1024,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,1024,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,768,0.006048000107208888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,768,0.005316444569163852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,256,0.0032426667296224167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,512,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,3584,0.009586667021115622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,2560,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,128,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,256,0.00564533347884814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,1024,0.004621333132187526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,7168,0.010374222364690568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,128,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,32,0.0028888889484935333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,512,0.003552000141806073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,64,0.00489688871635331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,512,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,65536,0.04865333437919617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,51200,0.029139555162853662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,64,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,51200,0.03871377640300327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,51200,0.044504887527889676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,65536,0.03635288940535651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,32,0.004968889057636261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,16384,0.016181333197487723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,65536,0.054752889606687755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,256,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,10240,0.010466666685210334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,128,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,16384,0.01314755529165268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,8192,0.01035911093155543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,10240,0.012259555359681448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,7168,0.009677333136399588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,8192,0.010999999940395355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,12288,0.01145244472556644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,6144,0.00905955582857132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,7168,0.010050666828950247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,7168,0.00959644466638565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,5120,0.007358222372002072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,12288,0.013457777599493662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,6144,0.008791111409664154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,4096,0.0073626670572492815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,5120,0.00904088881280687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,5120,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,10240,0.012253333297040729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,3584,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,4096,0.007694222033023834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,16384,0.01793333391348521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,3072,0.006331555545330048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,3584,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,2560,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,3072,0.007691555553012424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,3072,0.007002666592597961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,2048,0.006299555715587404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,2560,0.006611555400821898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,4096,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,2048,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,3584,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,1536,0.005965333431959152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,1024,0.004247111164861255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,1536,0.006813333266311222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,6144,0.009697777529557547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,768,0.003879111260175705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,1024,0.006567111031876669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,1024,0.005643555687533484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,512,0.0032515554792351196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,768,0.005963555640644497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,768,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,256,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,512,0.0052871112194326185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,8192,0.010017777482668558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,256,0.005984000033802456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,256,0.00490488898422983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,12288,0.014793778459231058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,64,0.0028764444092909494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,128,0.004964444372389051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,32,0.002569777684079276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,2048,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,512,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,32,0.004969777746333016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,65536,0.05473600162400139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,65536,0.046543111403783165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,128,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,2560,0.007355555891990662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,16384,0.014531556102964612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,51200,0.037581334511439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,1536,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,12288,0.012430222498046027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,128,0.0028560000161329904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,12288,0.013041777743233575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,10240,0.01145333300034205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,12288,0.013430222041077085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,10240,0.012100444071822695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,51200,0.043972445858849414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,8192,0.010799110763602786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,64,0.004906666775544484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,8192,0.010760889285140567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,7168,0.009704888694816166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,8192,0.009704888694816166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,6144,0.008719111482302347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,7168,0.010074666804737514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,7168,0.009349333743254343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,51200,0.02124799953566657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,6144,0.009347555538018545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,6144,0.00869511150651508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,4096,0.007324444750944774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,5120,0.008348444269763099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,10240,0.0110880004035102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,16384,0.01590222285853492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,4096,0.008177777959240807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,3584,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,3584,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,3584,0.0070186663005087115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,5120,0.008060444560315874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,3072,0.007622222105662028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,2560,0.005748444547255834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,2560,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,65536,0.025757332642873127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,2048,0.006325333482689328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,2560,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,3072,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,2048,0.0069431112044387394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,1536,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,2048,0.00619911112719112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,1024,0.004895110925038655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,5120,0.00868266655339135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,1536,0.006696888970004187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,768,0.00433688900536961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,1024,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,512,0.003551111039188173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,768,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,768,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,256,0.003208000005947219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,512,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,256,0.005659555395444234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,128,0.0028844444702068963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,3072,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,1536,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,1024,0.005588444570700328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,128,0.005616888817813661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,32,0.002579555536309878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,16384,0.015841777125994366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,64,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,32,0.004935111022657818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,256,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,768,51200,0.017816000514560275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,128,0.00499555551343494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,64,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,51200,0.03705777724583944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,768,16384,0.010397333237859938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,51200,0.04381777842839559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,768,12288,0.010801778071456485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,4096,0.007378666765160031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,768,65536,0.021703999903466966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,16384,0.014818666709793938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,768,10240,0.009682666924264696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,12288,0.013438222308953604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,768,8192,0.009334222310119206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,10240,0.012134222520722283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,65536,0.05413866705364651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,10240,0.011092444260915121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,8192,0.010762666662534079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,8192,0.009702222214804756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,768,6144,0.008310221963458592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,7168,0.010223111344708337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,7168,0.009359999663299983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,768,5120,0.0069848886794514125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,6144,0.009687110781669617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,512,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,6144,0.008725333544943068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,768,4096,0.007632888853549957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,5120,0.009001777403884465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,768,3584,0.007019555403126611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,5120,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,4096,0.007961778177155389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,768,3072,0.006863111009200414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,65536,0.04580444428655836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,768,7168,0.008663111262851292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,3584,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,3072,0.007631111476156447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,3072,0.006990222467316522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,768,2048,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,2560,0.007329777710967594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,2560,0.006678222368160884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,768,1536,0.005241777747869492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,2048,0.007000000112586551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,2048,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,768,1024,0.00423555572827657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,1536,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,4096,0.007347555624114142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,3584,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,1024,0.006335999816656113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,1024,0.0053191110491752625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,768,512,0.0034471110751231513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,768,0.006052444378534953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,768,0.005408000200986862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,768,256,0.0029191111938820947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,512,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,16384,0.015846222639083862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,768,128,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,256,0.0058062221441004015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,256,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,768,64,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,12288,0.012252444194422828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,768,32,0.0026382222357723448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,128,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,768,2560,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,64,0.00489777781897121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,512,65536,0.016906667086813185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,32,0.004908444566859139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,512,51200,0.01481599940194024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,65536,0.04829955432150099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,512,0.005026666654480828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,65536,0.04555911156866285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,512,16384,0.011037333144081963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,51200,0.038712888956069946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,512,12288,0.008674666285514832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,1536,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,128,0.005636444522274866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,12288,0.013455111119482251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,768,768,0.003923555629120933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,12288,0.012088889049159156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,512,8192,0.008343111309740279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,10240,0.012135110795497894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,10240,0.011056889262464313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,512,7168,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,8192,0.009711110757456886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,7168,0.010080888867378235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,51200,0.036772443188561335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,16384,0.01585244470172458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,16384,0.01482399966981676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,512,10240,0.007676444119877285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,6144,0.008784888519181145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,5120,0.009020444419648912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,5120,0.008050666915045844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,8192,0.010750222537252637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,4096,0.00816000004609426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,512,3584,0.007012444237867991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,512,6144,0.007332444190979004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,512,3072,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,3584,0.007962666451931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,3584,0.006979555719428592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,512,5120,0.006952889263629913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,3072,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,3072,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,512,2048,0.006108444597986009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,512,4096,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,2048,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,512,1536,0.004237333519591225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,4096,0.007640889121426477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,7168,0.009367999931176504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,512,1024,0.0038897775941424896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,1536,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,1536,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,6144,0.009371555513805812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,1024,0.0063777777055899305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,1024,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,512,512,0.0032622222271230486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,2560,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,768,0.0052391112678580815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,512,256,0.0028568889117903183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,512,0.005917333480384614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,512,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,2048,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,256,0.004613333278232151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,128,0.005616888817813661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,512,2560,0.006309333360857434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,768,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,512,32,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,2560,0.006437333507670297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,256,65536,0.0158942225906584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,512,128,0.0028764444092909494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,65536,0.04275200102064344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,65536,0.046055999067094594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,128,0.004592000196377437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,512,768,0.0035573335157500375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,256,16384,0.00888888869020674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,51200,0.03758666581577725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,51200,0.033943110042148165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,256,12288,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,64,0.004910222358173794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,16384,0.014291556345091926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,256,10240,0.007985778152942657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,12288,0.013153778182135688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,256,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,256,8192,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,10240,0.012096000214417776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,256,51200,0.013396444420019785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,256,7168,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,512,64,0.0025840000145965153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,8192,0.010033778018421596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,256,6144,0.006612444503439798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,7168,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,16384,0.015855999456511605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,256,5120,0.005961777849329843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,12288,0.012092444631788464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,6144,0.009358222285906473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,256,4096,0.005968889014588461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,5120,0.009000000026490953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,8192,0.010771555205186209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,256,3584,0.0063777777055899305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,7168,0.010055999788973067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,3584,0.007687110867765214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,3584,0.007038222418891058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,32,0.004934222333961063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,6144,0.008676444490750631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,5120,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,3072,0.00700711127784517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,256,2560,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,2560,0.007009777757856581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,256,2048,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,2048,0.005970666805903117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,2048,0.006792888873153263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,256,1536,0.004449777719047335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,10240,0.011002666420406766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,1536,0.0068764446510208984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,1536,0.005616888817813661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,4096,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,1024,0.006302222195598815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,256,768,0.0035333335399627686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,256,3072,0.006212444355090459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,256,512,0.0031866667171319327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,768,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,768,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,512,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,512,0.004934222333961063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,4096,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,256,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,256,128,0.002847111059559716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,256,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,256,64,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,128,0.005653333332803514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,256,32,0.002569777684079276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,128,0.004589333302444882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,128,65536,0.01424088908566369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,64,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,32,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,65536,0.04587999979654948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,256,256,0.0028684443483750024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,51200,0.0373617774910397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,51200,0.03388177686267429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,1024,0.005280888742870755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,16384,0.014489778214030795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,128,12288,0.007297777467303806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,16384,0.01585244470172458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,128,51200,0.012423999607563019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,12288,0.01314133322901196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,128,10240,0.007243555453088548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,10240,0.012078222301271228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,256,1024,0.0038933331767717996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,128,8192,0.006664000037643645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,8192,0.010779555473062726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,128,7168,0.00665155549844106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,8192,0.009707555174827576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,7168,0.010062221851613786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,128,6144,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,12288,0.012107555237081317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,6144,0.009401777552233802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,128,16384,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,128,5120,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,5120,0.009016888837019602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,128,4096,0.0057839999596277876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,3072,0.007650666766696506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,4096,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,128,3584,0.00601333338353369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,6144,0.00868444475862715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,3584,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,3584,0.007360888852013483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,5120,0.00798044436507755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,3072,0.007660444411966536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,128,2560,0.006224888894293044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,7168,0.009348444640636444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,2560,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,128,2048,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,10240,0.010667555862002902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,2048,0.006689777804745569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,2048,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,128,1536,0.004558222161398994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,1536,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,128,1024,0.003910222401221593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,1536,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,1024,0.006292444550328785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,128,768,0.0035342222286595237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,1024,0.0053102220926019884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,768,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,128,512,0.0031991110493739447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,768,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,128,3072,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,128,256,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,512,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,512,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,256,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,65536,0.041781332757737905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,256,0.004940444396601783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,128,64,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,128,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,128,32,0.0025128889828920364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,64,0.004936000125275718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,64,65536,0.01071466671095954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,32,0.004583111239804162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,64,51200,0.009703111317422655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,65536,0.04177600145339966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,3072,0.006690666493442323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,4096,0.007320889168315464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,51200,0.03351199958059523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,64,12288,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,2560,0.0063164445261160536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,64,10240,0.006299555715587404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,12288,0.011880888707107969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,10240,0.010759111079904767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,64,7168,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,128,128,0.0028737777223189673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,64,6144,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,128,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,64,5120,0.005988444305128521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,64,8192,0.006665777828958299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,64,4096,0.0057964444988303715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,5120,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,64,3584,0.006301333506902059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,4096,0.007535999847782983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,64,3072,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,64,16384,0.00701155596309238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,7168,0.009140444298585257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,3584,0.007196444604131911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,64,2048,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,3072,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,16384,0.014402666025691561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,64,1024,0.003934222377008862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,2048,0.006320888797442119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,64,768,0.0035413333939181436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,1024,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,64,512,0.0032195556494924757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,768,0.005311111195219888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,64,256,0.002862222285734283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,64,2560,0.006265777680608962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,64,128,0.002566222308410539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,512,0.0053004444473319584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,64,64,0.0025617778301239014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,64,1536,0.004460444466935264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,256,0.004955555415815777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,2560,0.006333333336644703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,6144,0.008303999900817871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,128,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,32,51200,0.008529778156015608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,64,0.004970666848950916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,32,0.0047022220161226065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,8192,0.009364444348547194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,32,12288,0.0063564446237352155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,51200,0.033942222595214844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,32,10240,0.006313777632183499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,16384,0.014467555615637036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,32,8192,0.005865777946180767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,12288,0.011931555966536203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,32,7168,0.005608888963858287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,64,32,0.002513777878549364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,32,6144,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,1536,0.00592533333433999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,7168,0.009033777647548253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,32,16384,0.006317333214812809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,32,4096,0.005588444570700328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,6144,0.008344000412358178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,32,3584,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,5120,0.007668444679843054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,32,3072,0.004978666702906291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,4096,0.007524444825119442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,32,2560,0.006065777606434292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,3584,0.0069822221994400024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,32,2048,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,3072,0.006637333167923822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,8192,0.009739555418491364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,2560,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,32,1024,0.003932444585694207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,2048,0.005986666513813867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,32,768,0.0034231110993358824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,65536,0.0414737794134352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,1536,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,32,512,0.0029102222373088202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,32,256,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,768,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,32,128,0.0028684443483750024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,10240,0.010825778047243753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,32,64,0.002543111021320025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,32,65536,0.010053333308961656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,128,0.004935111022657818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,32,5120,0.005921777751710679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,64,0.004734222259786394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,32,0.004591111093759537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,1024,0.005595555735958948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,512,0.0052675555149714155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,16384,0.3419066535101996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,51200,0.4860346582200792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,32,1536,0.004212444441186057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,7,32,32,0.002536000063021978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,51200,1.0199066797892253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,16384,0.2258995638953315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,12288,0.2612542311350505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,10240,0.20523999796973336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,12288,0.15263822343614367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,12288,0.10371200243631999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,8192,0.169995559586419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,10240,0.15015110704633924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,10240,0.08691733413272434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,8192,0.07151110967000325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,256,0.004602666530344221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,7168,0.06271021895938449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,7168,0.13312710656060114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,7168,0.0958168903986613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,6144,0.0888026687833998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,5120,0.09720089038213094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,6144,0.05503288904825846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,16384,0.13408888710869685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,4096,0.0805928905804952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,51200,0.43115999963548446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,5120,0.047288888030582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,3584,0.06892533434761895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,8192,0.11966578165690105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,4096,0.07048710849550036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,3072,0.059953775670793324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,6144,0.11486311753590901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,3584,0.04404977626270718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,2560,0.05197866757710775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,3072,0.04956266615125868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,2560,0.04350133405791389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,2048,0.04204622242185804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,4096,0.03962488969167074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,1536,0.032893333170149065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,2048,0.03599911265903049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,1536,0.02095288866096073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,1024,0.023732443650563557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,3072,0.036826666858461164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,1024,0.018233777748213876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,2560,0.02755822241306305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,768,0.018562666243977018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,768,0.015379556351237826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,2048,0.02310666607485877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,3584,0.061386664708455406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,256,0.008358221915033128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,512,0.012795555922720166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,512,0.013085333009560903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,1536,0.024298666252030268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,256,0.010770666930410596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,256,0.010013333625263637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,64,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,512,0.013437333206335703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,32,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,128,0.0063546668324205614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,32,0.00867733359336853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,128,0.010750222537252637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,5120,0.07306133376227485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,768,0.014478221535682678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,65536,0.5153422355651855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,65536,0.9697377946641711
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,51200,0.7501920064290365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,16384,0.25191466013590497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,51200,0.38889866405063206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,1024,0.020291555258962844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,12288,0.19270311461554634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,64,0.009230222139093611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,16384,0.14577333132425943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,12288,0.1229742235607571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,51200,0.35378310415479874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,10240,0.16163644525739881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,16384,0.108788443936242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,12288,0.08111733198165894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,10240,0.070214225186242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,7168,0.12467643949720596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,8192,0.09135911199781631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,8192,0.05749333567089505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,6144,0.10704622003767227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,7168,0.050366222858428955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,65536,0.44630665249294704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,6144,0.07293777995639376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,6144,0.046113779147466026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,4096,0.07400088840060763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,5120,0.047227554851108126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,8192,0.13298932711283365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,10240,0.13970843950907388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,5120,0.05922933419545492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,4096,0.033369776275422834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,4096,0.05443377627266777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,128,0.008001777860853408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,3584,0.029096000724368628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,3584,0.044293334086736046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,2560,0.04709866642951965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,7168,0.08256622155507405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,2048,0.03801244497299194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,5120,0.09078311257892185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,2560,0.02275111112329695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,1536,0.0298942228158315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,2048,0.019529778096410964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,3584,0.06329333119922213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,1536,0.02130399975511763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,1024,0.021607110897699993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,3072,0.055791113111707896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,768,0.016083555089102853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,1024,0.015823110938072205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,3072,0.025957332717047796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,512,0.011359111302428775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,768,0.013779555757840475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,2560,0.03570311268170675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,512,0.011720889144473605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,256,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,1536,0.017874666386180453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,128,0.0038977778620190094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,1024,0.017321777012613084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,3072,0.03950488898489211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,128,0.007339555356237624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,64,0.003544888976547453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,2048,0.030156443516413372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,32,0.0035617777870761026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,64,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,256,0.008346666892369589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,32,0.008010666403505538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,65536,0.32183289527893066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,768,0.012802666260136498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,51200,0.25013332896762425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,128,0.009701333112186855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,65536,0.15351466337839761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,16384,0.08711289034949408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,512,0.011168888873524137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,51200,0.13054222530788845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,12288,0.06637066602706909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,51200,0.12018577257792155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,16384,0.05099644594722324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,16384,0.06599555412928264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,10240,0.054640889167785645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,12288,0.033018665181265935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,12288,0.05099466774198744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,10240,0.02897600001758999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,7168,0.03710844450526767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,8192,0.024400000770886738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,65536,0.17289866341484916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,6144,0.03268266717592875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,7168,0.022256000174416438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,7168,0.03276000089115567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,5120,0.028388443920347426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,6144,0.021930666433440313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,6144,0.028619554307725694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,4096,0.024358222881952923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,256,0.009342222577995723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,5120,0.02216800053914388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,3584,0.02092711130777995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,8192,0.04263110955556234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,4096,0.022964444425370958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,3584,0.014490667316648694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,3584,0.018326222896575928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,8192,0.03671822283003066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,2560,0.015844444433848064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,3072,0.013478221992651621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,2048,0.013236444857385425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,2560,0.012416889270146688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,2560,0.01588355501492818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,1536,0.011042666931947073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,2048,0.013809777796268463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,5120,0.02510133385658264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,4096,0.01585422290696038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,3072,0.018486221631368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,10240,0.04560000035497877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,768,0.0069360000391801195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,1024,0.009016888837019602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,1024,0.009359999663299983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,512,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,768,0.008242666721343994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,768,0.0069857777820693115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,256,0.003528888854715559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,512,0.007485333416197035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,2048,0.011104889214038849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,512,0.006996444529957241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,256,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,256,0.006007110906971826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,1024,0.008348444269763099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,128,0.005585777676767773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,32,0.0028497777465316984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,64,0.005979555762476391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,3072,0.016837333639462788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,32,0.005969777703285217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,65536,0.24640978707207573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,128,0.0032417778339650896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,51200,0.18895110819074842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,65536,0.1389075517654419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,1536,0.010936000280910067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,16384,0.06658222277959187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,128,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,51200,0.11166132820977105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,12288,0.05152444375885857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,16384,0.04119822382926941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,16384,0.05582844548755222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,12288,0.02936089038848877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,12288,0.043169776598612465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,8192,0.037096000379986234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,10240,0.025609778033362493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,10240,0.03712266683578491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,7168,0.03235822253757053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,65536,0.14469156000349256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,8192,0.021760889225535925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,6144,0.02858577834235297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,8192,0.03169511093033685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,7168,0.020026667250527274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,5120,0.025164445241292317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,6144,0.019118222925398085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,51200,0.10593777894973755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,6144,0.025576000412305195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,5120,0.01904000010755327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,5120,0.020806221498383414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,3584,0.019668444991111755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,4096,0.014621333943472968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,4096,0.020265777905782063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,1536,0.01032266683048672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,3584,0.013280889226330651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,3584,0.016688000824716356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,7168,0.027679110566775005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,3072,0.012461333639091916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,2560,0.01536177761024899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,3072,0.015447111593352424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,10240,0.043952888912624784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,4096,0.022143110632896423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,2560,0.01149866647190518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,1536,0.008837333156002892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,2560,0.014813333749771118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,2048,0.010052444206343757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,1024,0.006956444846259222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,1536,0.009210666848553551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,1536,0.010103111465771994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,768,0.0058986664646201665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,3072,0.017551110850440133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,1024,0.008350222474998897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,512,0.003896000070704354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,768,0.007717333734035492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,768,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,512,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,512,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,128,0.0028479999552170434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,2048,0.012368000215954252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,128,0.006671999891599019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,128,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,64,0.002924444360865487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,32,0.0028728888266616394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,64,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,256,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,32,0.005924444645643234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,65536,0.20696443981594512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,256,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,51200,0.16227822833591038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,65536,0.13667199346754286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,65536,0.1323582198884752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,16384,0.058656891187032066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,64,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,12288,0.04414933257632785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,51200,0.10707733366224502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,16384,0.03607822126812405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,51200,0.10156889094246759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,10240,0.037999110089408025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,2048,0.013190222283204397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,12288,0.03851822349760268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,10240,0.024625778198242188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,256,0.00664444433318244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,8192,0.021026665965716045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,7168,0.028716444969177246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,1024,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,7168,0.019186667270130582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,6144,0.025729777084456548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,16384,0.04773155517048306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,5120,0.022097777989175584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,7168,0.023552000522613525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,6144,0.01869244376818339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,4096,0.019094222121768527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,10240,0.03154311246342129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,5120,0.019001776973406475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,3584,0.017437333861986797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,8192,0.026925333672099646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,4096,0.017237333787812125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,3584,0.012998222476906247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,3584,0.015152000718646579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,3072,0.012085333466529846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,3072,0.013213333156373767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,6144,0.02130399975511763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,8192,0.032144887579811945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,2048,0.01075466639465756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,2560,0.01242844429281023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,4096,0.015311999453438653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,1536,0.009003555609120263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,2048,0.009343110852771336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,3072,0.015828443898095023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,1024,0.006588444527652528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,1536,0.008801778157552084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,2560,0.014166222678290473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,768,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,2560,0.010620444185203977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,768,0.00757155567407608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,512,0.003572444534964032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,2048,0.011404444773991903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,512,0.006959111326270633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,256,0.0032391111469931076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,512,0.006330666856633292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,12288,0.027862221002578735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,128,0.0028862222615215513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,256,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,5120,0.015568888849682279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,64,0.0028400001012616684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,128,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,768,0.006301333506902059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,64,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,1536,0.00871377769443724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,32,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,65536,0.17239199744330513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,1024,0.008059555457697975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,1024,0.007693332930405934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,65536,0.07344177696439955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,51200,0.1295671065648397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,16384,0.04509511258867052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,51200,0.057137780719333224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,51200,0.06864444414774577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,12288,0.03324977888001336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,16384,0.023749333288934495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,16384,0.03471555643611484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,10240,0.02773333258099026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,12288,0.018574222922325134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,12288,0.028331554598278467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,8192,0.02401599950260586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,65536,0.08835111061731975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,10240,0.025191111697090998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,7168,0.02072177827358246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,8192,0.0151555554734336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,8192,0.020801777640978496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,32,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,6144,0.018179555733998615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,7168,0.014113777213626437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,5120,0.015819556183285184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,7168,0.018619555566045973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,6144,0.01277955538696713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,4096,0.013727111121018728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,5120,0.011744000017642975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,5120,0.015103111664454142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,3584,0.011888888974984487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,4096,0.013975999421543546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,128,0.005037333402368757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,3584,0.009724444813198512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,3072,0.011045333411958484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,2560,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,6144,0.017198221551047433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,3072,0.011409777734014722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,2048,0.008481777376598781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,256,0.006657777975002925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,2048,0.007820444802443186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,2048,0.009393778112199571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,3584,0.012493333054913415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,1536,0.007139555282062954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,1024,0.005658666706747479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,3072,0.008996444443861643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,2560,0.00831822223133511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,4096,0.011734222372372946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,1536,0.006996444529957241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,768,0.00609688874748018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,512,0.003638222399685118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,768,0.006690666493442323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,1536,0.007303111255168915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,256,0.0032248888164758682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,512,0.006480888773997624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,1024,0.006812444577614467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,512,0.0057502223385704895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,256,0.005304000029961268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,128,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,64,0.0028355556229750314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,64,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,32,0.002564444517095884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,10240,0.016334222422705758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,256,0.0061919999619325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,65536,0.14765955342186823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,128,0.005035555611054103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,65536,0.08336889081531101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,51200,0.062048892180124916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,2560,0.010495111346244812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,32,0.0053031109273433685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,51200,0.05363377928733826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,16384,0.03454666667514377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,65536,0.06645155615276761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,12288,0.02970577942000495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,768,0.0038968887594011095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,10240,0.02868888775507609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,10240,0.023580445183648005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,12288,0.03278399838341607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,8192,0.02437244521247016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,10240,0.015786666009161208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,8192,0.014132445057233175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,7168,0.018782221608691745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,8192,0.02033155494266086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,12288,0.017992888887723286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,6144,0.016709332664807636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,7168,0.01754399968518151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,16384,0.021576889687114294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,6144,0.012079111403889127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,5120,0.014427555931939019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,5120,0.011346666349305047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,5120,0.014166222678290473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,128,0.0028773333049482773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,16384,0.041182223293516375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,4096,0.012892444100644855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,3584,0.011391999820868174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,3584,0.008960888617568547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,3072,0.010079111489984725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,4096,0.013067555924256643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,1024,0.006728000111050076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,3072,0.008353778057628209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,3072,0.011044444309340583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,3584,0.011168888873524137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,2560,0.007684444387753804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,2560,0.010327999790509542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,51200,0.11751822630564372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,2048,0.008008889026112026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,1536,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,2560,0.008847999903890822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,4096,0.009923555784755284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,1024,0.004447999927732679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,1536,0.0069822221994400024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,2048,0.00797155582242542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,768,0.003834666477309333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,6144,0.016386666231685214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,1024,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,512,0.003220444545149803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,768,0.006325333482689328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,768,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,2048,0.008727110922336578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,512,0.006264888991912206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,512,0.005648889061477449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,128,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,256,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,7168,0.013144888811641268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,256,0.005252444495757421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,128,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,128,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,1024,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,65536,0.12234400378333198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,32,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,51200,0.09689511193169488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,65536,0.07481688923305936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,51200,0.05145066645410326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,16384,0.03556711143917508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,1536,0.00703288863102595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,16384,0.021060445242457922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,64,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,12288,0.028288000159793433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,12288,0.0173680004146364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,65536,0.06338844696680705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,12288,0.02748799986309475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,10240,0.01578044394652049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,51200,0.0561751127243042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,10240,0.021185777253574792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,32,0.0029057777590221832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,7168,0.019118222925398085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,8192,0.01824355622132619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,16384,0.02994133366478814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,6144,0.01699733402993944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,7168,0.016593777471118502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,256,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,10240,0.02441155579355028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,5120,0.015140444040298462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,6144,0.012187555432319641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,5120,0.010379555324713389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,6144,0.01556711064444648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,5120,0.01350844403107961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,3584,0.01036977767944336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,8192,0.013792888985739814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,4096,0.012625777886973487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,3584,0.008332444561852349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,3072,0.009341333475377824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,3584,0.010889777706729041
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,2560,0.008275555239783393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,3072,0.010168000227875179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,64,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,4096,0.011709333293967776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,4096,0.009360888765917884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,2048,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,1536,0.006308444258239534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,7168,0.012770666844315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,1536,0.00684533350997501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,1024,0.003963555726740095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,1536,0.006858666737874349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,8192,0.021551110678248938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,1024,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,768,0.0035911111368073356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,1024,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,512,0.0032026666320032547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,768,0.006292444550328785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,768,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,2048,0.007296000089910295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,512,0.005571555760171678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,3072,0.008343111309740279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,256,0.0052888890107472735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,128,0.0028737777223189673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,128,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,2560,0.007435555259386699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,64,0.002625777696569761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,128,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,32,0.0025724443710512584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,65536,0.10642400052812363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,64,0.00526577772365676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,512,0.006269333263238271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,51200,0.0843120018641154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,65536,0.061117331186930336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,256,0.005961777849329843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,16384,0.03170311119821336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,51200,0.04933599962128533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,2560,0.009740444521109263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,12288,0.02574399941497379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,16384,0.020271999968422782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,256,0.0032017777363459268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,12288,0.02238311039076911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,10240,0.02180533276663886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,65536,0.0672560003068712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,12288,0.0170133329100079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,8192,0.019820445113711886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,10240,0.020078221956888836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,51200,0.05087466703520881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,7168,0.0174666659699546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,8192,0.013443555268976422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,16384,0.02886399957868788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,8192,0.017256889078352187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,7168,0.012490666574902005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,7168,0.015506666567590503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,32,0.004978666702906291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,6144,0.011439111497667102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,6144,0.014744000302420722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,4096,0.011060444845093621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,5120,0.010072000324726105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,5120,0.012481778032249875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,3584,0.009448889229032729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,4096,0.00869333330127928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,4096,0.012138666378127204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,3072,0.00868977771864997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,3584,0.010080888867378235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,6144,0.015672889020707872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,3072,0.007641777396202087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,3072,0.00941955546538035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,2048,0.006822222222884496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,2560,0.00868000007337994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,2048,0.00720355576939053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,10240,0.014875556031862894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,2048,0.007296000089910295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,3584,0.008366222182909647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,1024,0.003940444439649582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,1536,0.0069439998931354946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,1024,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,768,0.003530666646030214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,2560,0.007320889168315464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,768,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,512,0.003232888877391815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,768,0.005647111270162795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,512,0.006136888845099344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,256,0.003007111036115222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,512,0.0052675555149714155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,1536,0.00674933319290479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,2560,0.007756444315115611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,256,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,5120,0.01387288918097814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,64,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,128,0.004964444372389051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,32,0.002534222271707323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,64,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,65536,0.09045155843098958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,128,0.0028657778683635923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,65536,0.05840088923772176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,256,0.0052328887912962176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,1024,0.006333333336644703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,51200,0.06640177965164185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,16384,0.026463111241658527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,51200,0.04762399858898587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,51200,0.046219554212358266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,1536,0.00543200017677413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,12288,0.020671111014154222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,16384,0.025847999585999384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,10240,0.018577777677112155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,12288,0.021312889125612047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,10240,0.014506666196717156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,8192,0.01608444419172075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,128,0.005919111271699269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,8192,0.012879110872745514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,7168,0.014383110735151501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,2048,0.007686222592989604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,16384,0.019650666250122916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,7168,0.011789333489206104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,7168,0.01516177753607432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,6144,0.010384889112578498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,5120,0.01197866681549284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,65536,0.05894222524431017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,5120,0.00904266701804267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,5120,0.012200000385443369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,4096,0.010826666322019366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,4096,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,3584,0.011041777829329172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,4096,0.011407111254003314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,3584,0.008034666379292807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,3072,0.01036800030204985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,10240,0.019519999623298645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,3072,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,2560,0.008925333619117737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,8192,0.01625955601533254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,2560,0.007322666545708974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,2560,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,6144,0.01346577786737018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,2048,0.007074666519959767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,1536,0.005963555640644497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,3584,0.00869066682126787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,6144,0.013843555417325763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,1024,0.004960888789759742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,1536,0.006660444455014334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,1536,0.00666133314371109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,1024,0.005971555494599872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,1024,0.0063182223174307085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,768,0.004568888909286923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,32,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,2048,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,768,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,768,0.005659555395444234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,256,0.0031840000301599503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,2048,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,256,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,256,0.005137777576843898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,3072,0.008354666332403818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,128,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,128,0.004963555683692296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,512,0.0038817777401871155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,32,0.0025680000997251934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,512,0.006005333529578314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,512,0.005283555636803309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,65536,0.07792800002627902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,128,0.0029048888633648553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,65536,0.0580293337504069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,64,0.002540444541308615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,12288,0.016176889340082806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,51200,0.044532444741990834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,16384,0.02439555525779724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,51200,0.04685689012209574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,64,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,16384,0.019289778338538278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,10240,0.016467556357383728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,12288,0.01607911123169793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,12288,0.02084000077512529
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,8192,0.014761777387724983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,10240,0.014512888259357877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,65536,0.05656355619430542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,10240,0.018897778458065458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,7168,0.013614222407341003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,8192,0.012442666623327466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,8192,0.016187555260128446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,7168,0.011408888631396823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,7168,0.013839999834696451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,5120,0.011386666860845355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,6144,0.010011555420027839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,12288,0.018901333212852478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,6144,0.013384888569513956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,4096,0.010229333407349056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,32,0.004910222358173794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,3584,0.01090311093462838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,4096,0.008444444172912175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,4096,0.009712888962692684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,51200,0.06081866555743747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,3584,0.008003555238246918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,3584,0.0087333329849773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,2560,0.009008000294367472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,3072,0.007654222349325816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,3072,0.007592889169851939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,2560,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,2560,0.007679999702506595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,16384,0.025103999508751765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,2048,0.0070186663005087115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,3072,0.009352000223265754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,1536,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,1024,0.004907555464241239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,6144,0.012436444560686747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,1024,0.006400888992680444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,768,0.00425511101881663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,1024,0.0058222222659322954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,5120,0.009431111315886179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,1536,0.006268444574541516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,768,0.005670222143332164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,512,0.005975111077229182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,256,0.0032257777121331957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,1536,0.006608888920810487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,256,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,2048,0.007290666302045186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,2048,0.00703022215101454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,5120,0.012797333300113678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,64,0.002552888873550627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,128,0.0059057776298787855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,32,0.002874666617976295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,512,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,64,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,65536,0.06343199809392293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,32,0.004949333353175057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,51200,0.050476445092095264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,128,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,65536,0.05717955695258247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,768,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,51200,0.045935998360315956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,16384,0.01926577753490872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,12288,0.017214222086800467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,12288,0.019478221734364826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,256,0.004971555537647671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,65536,0.05425688955518934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,12288,0.01585422290696038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,16384,0.02057422200838725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,128,0.004912000149488449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,10240,0.014096889230940076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,10240,0.016526222229003906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,7168,0.012400888734393649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,8192,0.012072888513406118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,16384,0.022376888328128394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,8192,0.014807111687130399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,6144,0.011391110718250275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,7168,0.010393777655230628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,7168,0.01315822203954061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,5120,0.010404444403118556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,6144,0.012088889049159156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,5120,0.009120889008045197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,512,0.003547555456558863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,4096,0.008728000024954477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,4096,0.008353778057628209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,3584,0.008914666871229807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,6144,0.009720000127951304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,10240,0.015188443991872998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,3072,0.009702222214804756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,3584,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,3584,0.008295999632941352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,51200,0.04280622137917412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,3072,0.007636444436179266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,3072,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,2560,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,2560,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,2048,0.007096000015735626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,1536,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,1536,0.006261333409282897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,8192,0.013063111239009433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,2048,0.00666222224632899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,1024,0.005697777701748743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,1024,0.0064026667839950984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,768,0.004597333156400257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,2048,0.006607111129495833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,512,0.003882666842805015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,768,0.005588444570700328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,512,0.005983110931184556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,5120,0.010371555884679159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,256,0.003228444399105178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,256,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,128,0.0032417778339650896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,256,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,128,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,768,0.006231110956933763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,64,0.0028782222006056043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,32,0.0028782222006056043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,1536,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,64,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,65536,0.05684355232450697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,32,0.004940444396601783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,51200,0.04465599854787191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,65536,0.05610666672388712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,65536,0.05254488852288988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,16384,0.018732445107565988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,51200,0.04526133338610331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,51200,0.04122844338417053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,128,0.00490488898422983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,16384,0.018903111418088276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,16384,0.021432888176706102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,12288,0.01554844445652432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,12288,0.017418666018380057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,10240,0.013413333230548434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,8192,0.011973333027627734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,8192,0.013691555294725629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,12288,0.01531022290388743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,7168,0.011387555963463254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,8192,0.011121778024567498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,10240,0.013691555294725629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,7168,0.010412444670995077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,6144,0.010738666686746808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,7168,0.012098666694429187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,5120,0.009742221898502773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,6144,0.009709333380063375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,6144,0.011025778121418424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,4096,0.009053332938088311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,5120,0.009026666482289633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,5120,0.008683555656009251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,3584,0.00867022242810991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,4096,0.008391111261314815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,4096,0.008055999875068665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,3072,0.008010666403505538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,3584,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,512,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,4096,0.009415110780133141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,3072,0.007650666766696506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,3072,0.007130666739410824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,2048,0.005903111149867375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,2560,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,2560,0.006642666541867786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,1536,0.005903999838564131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,2048,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,2048,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,1024,0.004602666530344221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,1536,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,3584,0.008030222521887885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,768,0.0042462220622433555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,2560,0.007299555672539606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,768,0.006119110931952794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,10240,0.015949333707491558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,512,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,1536,0.006643555644485686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,256,0.0032231110251612137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,512,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,1024,0.006351999938488007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,1024,0.004869333157936732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,128,0.0031999999450312722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,256,0.005783111270931032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,512,0.003544888976547453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,128,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,128,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,32,0.002522666628162066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,64,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,65536,0.0500257776843177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,1024,0.00582755563987626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,256,0.00490133340160052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,51200,0.03811911079618666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,768,0.005626666463083691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,64,0.0028568889117903183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,16384,0.01590044465329912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,51200,0.04481066597832573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,51200,0.03997244437535604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,12288,0.012897777888509961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,16384,0.01848355597919888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,32,0.004908444566859139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,10240,0.011791110866599612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,12288,0.014359111587206522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,2560,0.007299555672539606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,8192,0.011051555474599203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,65536,0.049977779388427734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,10240,0.013770666387346057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,65536,0.05552800165282356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,7168,0.010482666393121084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,8192,0.011101333631409539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,6144,0.009680888719028896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,8192,0.01201866649919086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,7168,0.010377777947319878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,7168,0.010642666783597736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,6144,0.009583111438486312
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,4096,0.010034666293197209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,16384,0.02034311162100898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,5120,0.009095999929640029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,5120,0.00828888929552502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,3584,0.00980444418059455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,4096,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,12288,0.01573066744539473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,4096,0.007663999994595845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,10240,0.01240888900227017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,2560,0.005983110931184556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,3072,0.006955555743641323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,5120,0.009000889129108852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,2560,0.006682666639486949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,6144,0.008943999807039896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,2048,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,1536,0.005671999934646819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,3584,0.008185777399275038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,3584,0.007304888632562425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,3072,0.007636444436179266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,1536,0.005991999887757831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,2560,0.007294221884674496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,1024,0.006333333336644703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,768,0.004222222086456087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,768,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,512,0.003532444437344869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,768,0.00565244464410676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,2048,0.006980444822046492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,3072,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,512,0.005980444451173146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,512,0.0052266667286554975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,1024,0.004622222234805425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,128,0.002847111059559716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,256,0.004951111144489712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,64,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,128,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,1024,0.005420444326268301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,64,0.004935111022657818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,65536,0.035996443695492215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,32,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,256,0.0032151111712058387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,51200,0.028571556011835735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,65536,0.04849511053827074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,256,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,16384,0.01328888866636488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,51200,0.04451555675930447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,32,0.0026613333159022858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,16384,0.01789066692193349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,16384,0.01761244402991401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,12288,0.013432000246312885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,12288,0.01458755632241567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,10240,0.010667555862002902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,65536,0.05492000116242302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,8192,0.010249777800507015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,10240,0.013462222284740873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,51200,0.038691557115978666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,8192,0.010866666833559671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,8192,0.010049777726332346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,12288,0.011386666860845355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,6144,0.009013333254390294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,7168,0.010079999764760336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,7168,0.009373333719041612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,5120,0.007679111427730984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,6144,0.009701333112186855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,6144,0.009025777379671732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,5120,0.008837333156002892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,10240,0.012148444851239523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,5120,0.008054222497675154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,4096,0.008340444829728868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,7168,0.009745777481132084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,2048,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,3584,0.008000888758235509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,3072,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,3584,0.007291555404663086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,3072,0.007660444411966536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,2560,0.005636444522274866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,3072,0.0069546666410234236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,1536,0.006807999892367258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,2048,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,2560,0.007290666302045186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,2560,0.006443555570311016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,3584,0.007028444773621029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,4096,0.007690666450394525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,1536,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,1024,0.004229333251714706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,1024,0.005655111124118169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,1024,0.0063688887490166565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,768,0.003936000168323517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,2048,0.006980444822046492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,512,0.003516444315512975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,768,0.006236444330877728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,2048,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,256,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,512,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,512,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,128,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,256,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,256,0.004980444494220945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,64,0.0028293333533737394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,128,0.005655111124118169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,128,0.004910222358173794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,1536,0.005250666704442766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,128,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,64,0.004909333255555895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,32,0.004941333499219683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,51200,0.021037333541446265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,768,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,65536,0.04665155543221367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,16384,0.014813333749771118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,51200,0.043895999590555825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,51200,0.03745066788461473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,4096,0.007288888924651676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,12288,0.01242844429281023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,16384,0.0158906661801868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,16384,0.01591199967596266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,10240,0.011156444748242697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,12288,0.013412444127930535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,8192,0.01036622209681405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,12288,0.013264888690577613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,10240,0.012082666986518435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,7168,0.009711110757456886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,8192,0.010727110836240979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,65536,0.054689778221978076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,8192,0.009696000152164036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,6144,0.00923111124171151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,7168,0.010082667072614035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,5120,0.008391111261314815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,6144,0.009699555734793345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,6144,0.00871733327706655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,4096,0.007634667058785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,5120,0.008050666915045844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,32,0.0025759999536805684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,3584,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,4096,0.008045333127180735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,65536,0.025973333252800837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,10240,0.01107377807299296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,3584,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,3584,0.007268444531493717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,2560,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,3072,0.007346666521496243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,3072,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,2048,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,2560,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,5120,0.00906488878859414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,1536,0.005624889085690181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,2048,0.006981333096822103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,1536,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,3072,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,1536,0.0057964444988303715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,7168,0.009436444275909001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,2560,0.00701333334048589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,2048,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,4096,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,1024,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,512,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,768,0.00416355538699362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,256,0.0032097777972618737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,512,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,256,0.00573600000805325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,768,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,128,0.003097777772280905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,256,0.005048000150256687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,768,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,64,0.0025617778301239014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,128,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,32,0.002891555635465516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,128,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,768,65536,0.021316443880399067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,64,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,1024,0.005630222045713001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,32,0.004982222285535601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,65536,0.05407911207940844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,65536,0.04581955406400892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,768,16384,0.01037688884470198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,51200,0.04359822140799629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,51200,0.03707111212942336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,768,12288,0.010799110763602786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,16384,0.014612444572978549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,512,0.003543111185232798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,768,10240,0.010031999813185798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,12288,0.013175999952687157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,12288,0.012402666939629449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,768,8192,0.00906488878859414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,10240,0.012030222349696688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,1024,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,768,7168,0.009033777647548253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,8192,0.01071999967098236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,8192,0.010050666828950247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,7168,0.010035555395815108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,768,6144,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,7168,0.009203555683294931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,16384,0.015837333268589444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,6144,0.009610666996902889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,1536,0.005990222096443176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,6144,0.00868444475862715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,5120,0.008954666554927826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,5120,0.008023999631404877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,768,3584,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,4096,0.008387555678685507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,768,51200,0.017768889665603638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,768,3072,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,3584,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,3584,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,768,2560,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,3072,0.006981333096822103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,768,5120,0.007019555403126611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,2560,0.007293333609898885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,2560,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,768,4096,0.0073626670572492815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,2048,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,2048,0.00612088872326745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,768,1024,0.003912888881233004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,1536,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,4096,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,768,768,0.003543111185232798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,1024,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,1024,0.005320000151793162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,768,512,0.003265777809752358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,768,0.0052791109515561005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,768,2048,0.005240889059172736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,768,256,0.0032071111102898917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,512,0.005116444494989184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,256,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,256,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,10240,0.011066666907734342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,128,0.005586666779385672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,768,64,0.0026782221264309357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,768,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,768,32,0.002895999906791581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,512,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,768,1536,0.00526577772365676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,32,0.004912000149488449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,768,128,0.002874666617976295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,512,51200,0.014708444476127625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,128,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,3072,0.007699555820888943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,512,16384,0.010821333361996545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,51200,0.03866488734881083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,64,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,16384,0.015855110353893705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,16384,0.014513777361975776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,65536,0.04996711015701294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,512,10240,0.007660444411966536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,10240,0.011069333387745751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,10240,0.012077333198653327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,512,65536,0.016742222838931613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,512,8192,0.008341333104504479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,8192,0.010799999866220685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,512,7168,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,7168,0.009362666971153682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,51200,0.03646399908595615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,512,6144,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,512,12288,0.008671110702885522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,6144,0.00869422240389718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,512,5120,0.006994666324721442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,1536,0.006224000205596288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,5120,0.008955555657545725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,5120,0.007663999994595845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,7168,0.010040889183680216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,4096,0.006979555719428592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,512,3584,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,4096,0.00828888929552502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,6144,0.009642666412724389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,3584,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,65536,0.045365333557128906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,3072,0.00664533343580034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,3072,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,512,2560,0.006506666541099548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,2560,0.006312888943486744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,512,3072,0.006923555499977536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,12288,0.01354311075475481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,2560,0.007330666813585494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,2048,0.006974221931563483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,512,1536,0.004581333448489507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,1536,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,1536,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,512,1024,0.004129777765936322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,12288,0.012077333198653327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,512,768,0.003590222034189436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,1024,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,512,4096,0.007337777978844113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,512,512,0.0032151111712058387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,768,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,512,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,512,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,8192,0.009708444277445475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,256,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,512,128,0.0025600000388092464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,2048,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,128,0.005708444449636672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,128,0.004635555462704765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,512,256,0.003080888961752256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,512,32,0.0025680000997251934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,64,0.004638222356637319
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,512,2048,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,256,65536,0.014803555276658801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,32,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,256,51200,0.013458666702111563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,65536,0.04622755448023478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,768,0.006085333310895496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,256,16384,0.008997333546479544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,51200,0.03734933336575826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,51200,0.034627556800842285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,512,64,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,256,12288,0.0083155557513237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,16384,0.01585244470172458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,256,10240,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,12288,0.01315199997689989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,12288,0.012321777641773224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,3584,0.007038222418891058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,10240,0.012081777883900536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,10240,0.010788444843557147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,256,7168,0.006618666566080517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,8192,0.010762666662534079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,8192,0.009731555150614845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,256,6144,0.006312888943486744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,65536,0.04241333405176798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,7168,0.009073778159088558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,256,5120,0.005922666854328579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,6144,0.009687110781669617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,16384,0.014490667316648694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,256,4096,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,5120,0.00798044436507755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,256,0.004951999833186467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,256,8192,0.007027555671003129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,256,3584,0.006325333482689328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,4096,0.008004444340864817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,4096,0.0073724447025193115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,256,3072,0.005923555543025334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,3584,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,256,2560,0.0063440000845326324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,3584,0.007045333584149678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,3072,0.0069848886794514125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,256,2048,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,2560,0.006329777754015393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,6144,0.00867733359336853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,256,1536,0.00462311092350218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,2048,0.0069955554273393415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,2048,0.006339555399285422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,256,1024,0.003567111161020067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,1536,0.006666666517655055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,1536,0.005656888915432825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,1024,0.005626666463083691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,256,768,0.003540444291300244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,1024,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,3072,0.007708444363541073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,768,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,768,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,2560,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,256,256,0.002874666617976295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,512,0.004936000125275718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,256,128,0.002524444419476721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,256,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,256,0.004997333304749595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,256,64,0.002554666664865282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,128,0.005618666609128316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,256,32,0.0026640000028742682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,128,0.004655111167165968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,1024,0.006337777607970768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,64,0.005034666508436203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,7168,0.010057777994208867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,32,0.004908444566859139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,128,51200,0.01276088919904497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,65536,0.04625066783693102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,128,16384,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,51200,0.03737422161632114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,5120,0.009005332986513773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,51200,0.033839109871122576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,128,12288,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,16384,0.015834665960735746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,16384,0.014467555615637036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,12288,0.01314222233162986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,128,65536,0.013968888256284924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,256,512,0.0032008888406885993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,10240,0.011733333269755045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,128,8192,0.0070222218831380205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,10240,0.011117333339320289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,65536,0.042103111743927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,8192,0.0107360002067354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,8192,0.009709333380063375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,128,6144,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,7168,0.009391999906963771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,6144,0.00941422250535753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,128,10240,0.007287999822033777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,6144,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,5120,0.008671110702885522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,128,4096,0.00565422202150027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,512,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,4096,0.007984889050324758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,128,3584,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,3584,0.00703288863102595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,12288,0.012094222009181976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,128,3072,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,3584,0.007691555553012424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,128,2560,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,5120,0.008027555214034187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,128,7168,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,128,2048,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,2560,0.007291555404663086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,2560,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,4096,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,2048,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,2048,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,128,1024,0.0038808890514903595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,1536,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,1024,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,1024,0.005592888842026393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,3072,0.007678222325113084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,3072,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,768,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,128,512,0.003206222214632564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,7168,0.010055111514197456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,128,256,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,512,0.005969777703285217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,1536,0.006614222294754452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,512,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,128,128,0.0026346666531430352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,128,5120,0.006263999889294307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,128,64,0.002524444419476721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,128,0.005706666658322017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,128,32,0.002518222149875429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,128,768,0.003552888830502828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,64,65536,0.010372444159454769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,64,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,64,51200,0.010024000373151567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,32,0.004940444396601783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,64,16384,0.007082666787836287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,65536,0.04144355654716492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,64,12288,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,51200,0.033922665648990206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,64,10240,0.005975111077229182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,16384,0.014302222265137566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,256,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,12288,0.011921777493423887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,64,7168,0.006229333165619109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,128,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,64,6144,0.005929777605666055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,8192,0.009701333112186855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,64,5120,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,768,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,7168,0.009031111167536842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,64,4096,0.005631999837027655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,128,1536,0.004249777644872665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,6144,0.008654221892356873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,64,3072,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,5120,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,64,2560,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,3584,0.006981333096822103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,64,2048,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,2560,0.006301333506902059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,10240,0.011060444845093621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,64,1536,0.004564444637960858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,64,1024,0.003910222401221593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,2048,0.005975111077229182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,64,768,0.0032008888406885993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,1536,0.005972444597217772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,64,3584,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,768,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,64,256,0.0025884444928831523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,256,0.005623999983072281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,512,0.005290666802061929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,3072,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,64,64,0.0025600000388092464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,256,0.004607999904288186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,64,32,0.002528000002106031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,128,0.005020444591840108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,64,0.00489777781897121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,32,65536,0.009386666946940953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,64,512,0.0031822222388452957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,32,0.005020444591840108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,4096,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,32,16384,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,64,128,0.002817777709828483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,51200,0.03399111164940728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,32,12288,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,32,10240,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,16384,0.014125333891974555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,32,8192,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,12288,0.012123555772834353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,32,7168,0.0063235556913746735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,1024,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,8192,0.009680888719028896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,7168,0.00904888908068339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,32,5120,0.005991999887757831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,64,8192,0.006504888749784893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,6144,0.008559111091825698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,5120,0.008018666671382057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,32,3584,0.005706666658322017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,4096,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,32,3072,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,32,51200,0.008384000096056197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,32,2560,0.005976000179847081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,3072,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,32,2048,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,32,1536,0.004396444393528832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,32,4096,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,32,1024,0.0035946667194366455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,2048,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,32,768,0.003543111185232798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,10240,0.01072799993885888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,32,512,0.0031893334041039148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,768,0.005286222116814719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,32,6144,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,32,256,0.002867555452717675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,32,128,0.0028408887899584244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,65536,0.04180888997183906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,512,0.004971555537647671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,256,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,2560,0.006313777632183499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,128,0.004587555511130227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,64,0.0046471113132105935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,1536,0.005623111294375525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,32,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,1024,0.005607999861240387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,3584,0.0070159998204973005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,16384,0.3390524387359619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,51200,0.48632086647881406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,16384,0.2461315525902642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,32,32,0.002514666567246119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,16384,0.13385778003268772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,12288,0.18386667304568818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,12288,0.10293066501617432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,51200,0.4365982214609782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,8192,0.1685262256198459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,10240,0.08654222223493789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,6,32,64,0.002513777878549364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,7168,0.14760355154673258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,51200,1.011880874633789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,8192,0.07134044170379639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,12288,0.2603742281595866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,6144,0.11509422461191814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,7168,0.10124888685014512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,10240,0.21840533945295545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,6144,0.06871289014816284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,4096,0.08027555545171101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,5120,0.09703999757766724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,10240,0.1244000064002143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,3584,0.06990310880872938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,4096,0.03905333413018121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,4096,0.06414755847718981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,8192,0.11806044313642715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,3584,0.03499999973509047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,3584,0.0510755545563168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,2560,0.05055200060208639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,7168,0.06251111295488146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,3072,0.031238221459918555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,6144,0.09872355726030137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,2048,0.04189688960711161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,2560,0.027383999692069158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,5120,0.0469413333468967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,1536,0.0325546662012736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,2048,0.02327022287580702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,2048,0.034267554680506386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,1536,0.022664889693260193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,1536,0.027440889014138117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,768,0.018803555104467604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,3072,0.06092088752322727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,768,0.015375110838148328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,3072,0.05169066786766052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,5120,0.07667999797397189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,2560,0.039297779401143394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,1024,0.023444443941116333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,256,0.008295111358165741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,256,0.010768888725174798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,128,0.006256888724035687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,768,0.01681066718366411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,512,0.012934221989578672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,512,0.012944888737466602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,64,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,1024,0.01789688898457421
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,32,0.005625777774386936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,64,0.0090524446633127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,512,0.013574221895800697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,128,0.008000888758235509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,128,0.010617777705192566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,65536,0.5117413202921549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,65536,0.9701271057128906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,16384,0.2506382200453016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,51200,0.7477902306450738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,32,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,51200,0.3884062237209744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,12288,0.1921164459652371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,1024,0.020337777005301583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,51200,0.3503742218017578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,16384,0.19468888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,10240,0.16133333577050105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,16384,0.110225776831309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,8192,0.13267021709018284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,10240,0.06963911321428087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,10240,0.09909244378407796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,8192,0.09206310907999675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,7168,0.1166622241338094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,8192,0.05979377693600125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,6144,0.10631022188398574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,256,0.01033333357837465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,5120,0.09045510821872288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,12288,0.11915911568535699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,6144,0.0780186653137207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,12288,0.08312533299128215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,4096,0.07405777772267659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,5120,0.06085599793328179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,3584,0.06313600142796834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,4096,0.03321155574586656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,65536,0.4247599972618951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,3584,0.029659556017981634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,3584,0.041101333167817854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,7168,0.05158488949139913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,3072,0.05582133266660902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,3072,0.026452443665928308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,6144,0.04474222328927782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,2048,0.038014223178227745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,5120,0.04187466700871786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,2048,0.019684443871180218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,1536,0.029498666524887085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,7168,0.08178577820460002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,2560,0.04685599936379326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,1024,0.02119111186928219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,1536,0.019741333193249173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,3072,0.04240533378389147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,768,0.01623288955953386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,2560,0.02310666607485877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,1024,0.016985777351591323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,512,0.011408888631396823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,4096,0.05427289009094238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,768,0.014478221535682678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,256,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,512,0.011408000356621213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,512,0.011399110986126794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,128,0.0042204442951414315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,1536,0.02367288867632548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,256,0.008663111262851292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,64,0.003522666792074839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,128,0.009676444033781687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,32,0.0035955554081334006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,2560,0.03268622358640035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,64,0.008191111187140146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,32,0.007737778127193451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,2048,0.028110222684012517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,51200,0.2485724555121528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,256,0.009703999592198266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,65536,0.1523724397023519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,128,0.007335111498832703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,1024,0.015604444675975375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,51200,0.12307111422220866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,768,0.013792888985739814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,12288,0.06679199801550971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,16384,0.04190311166975233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,65536,0.3212035497029622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,10240,0.0548000004556444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,12288,0.033372445238961115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,12288,0.050350222322675914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,8192,0.04336444536844889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,65536,0.17218132813771567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,7168,0.036646223730511136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,10240,0.04258844587537977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,8192,0.02442933287885454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,8192,0.03682044479582045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,16384,0.0867991116311815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,7168,0.022076444493399724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,7168,0.03271911210483975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,6144,0.02544266647762722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,6144,0.028382221857706707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,16384,0.060468441910213895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,4096,0.024163555767801072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,5120,0.01872711049185859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,5120,0.024944000773959692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,3584,0.020595555504163105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,4096,0.021537777450349595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,3584,0.01610666679011451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,3584,0.018803555104467604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,51200,0.13052444987826878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,5120,0.027890665663613215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,2560,0.01587022178702884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,3072,0.018000889155599806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,2560,0.01240711079703437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,2560,0.014816888504558139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,4096,0.01571466690964169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,10240,0.028596444262398615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,1536,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,2048,0.011062222222487131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,2048,0.013770666387346057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,3072,0.01850133306450314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,1024,0.008267555799749162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,1536,0.011906666888131036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,1024,0.009349333743254343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,1024,0.009053332938088311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,768,0.008220444122950235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,768,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,2048,0.013077333569526672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,512,0.007648000286685095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,256,0.003237333355678452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,6144,0.03260622090763516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,1536,0.010891555911964841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,128,0.0031759999692440033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,256,0.006959999601046245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,768,0.0069466663731469055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,64,0.0028417776856157514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,128,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,128,0.005581333405441708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,512,0.0046880000995265115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,65536,0.24283289909362793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,64,0.006020444548792309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,512,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,51200,0.1886542240778605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,65536,0.14710222350226507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,16384,0.0661733349164327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,65536,0.13620711697472468
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,3072,0.013442666994200813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,51200,0.11167289151085748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,12288,0.051579554875691734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,51200,0.10696444246504043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,16384,0.03772711091571384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,10240,0.0439591109752655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,12288,0.029622223642137315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,12288,0.04400711258252462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,32,0.005923555543025334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,10240,0.025736888249715168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,7168,0.0325582226117452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,8192,0.021990221407678392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,8192,0.03250222073660956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,6144,0.028853333658642236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,7168,0.01998933321899838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,7168,0.027929776244693335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,5120,0.025781333446502686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,32,0.0029084444459941653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,6144,0.025164445241292317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,16384,0.052183998955620654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,4096,0.022088888618681166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,5120,0.01700888905260298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,5120,0.022862222459581163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,3584,0.01942488882276747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,4096,0.014647111296653748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,4096,0.019926221834288705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,3584,0.013545778062608508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,3584,0.015984889533784654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,3072,0.012562666502263812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,3072,0.016528889536857605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,256,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,2048,0.013119999733236102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,2560,0.013529777526855469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,8192,0.036743111080593534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,1536,0.008635555704434713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,2048,0.012255111502276527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,3072,0.017281777328915067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,1024,0.007224889265166388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,1536,0.01071999967098236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,2560,0.01553600033124288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,768,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,1024,0.00848711116446389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,6144,0.018213333355055917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,512,0.0038764443662431505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,768,0.007641777396202087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,768,0.00741511086622874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,512,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,512,0.006653333289755716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,1536,0.00981866651111179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,256,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,128,0.0028924445311228433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,128,0.0069875551594628235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,1024,0.008357333640257517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,128,0.005474666754404704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,2560,0.011342222491900126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,64,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,10240,0.036995556619432234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,65536,0.20491022533840605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,256,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,65536,0.14147822062174478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,64,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,65536,0.1272897720336914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,32,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,16384,0.058267553647359215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,256,0.0031884445084465873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,51200,0.10205155611038208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,32,0.005708444449636672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,12288,0.043924444251590304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,16384,0.050604446066750415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,10240,0.037696000602510236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,12288,0.02869422237078349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,12288,0.036900444163216486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,8192,0.031903998719321355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,10240,0.02499733368555705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,10240,0.031168000565634832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,7168,0.028453333510292902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,8192,0.02126577827665541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,8192,0.026485333840052288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,6144,0.025616000096003216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,7168,0.019294222195943195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,7168,0.023174222972657945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,5120,0.022035555707083807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,16384,0.035954667462242976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,6144,0.020224889119466145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,4096,0.019262222780121695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,5120,0.018570666511853535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,51200,0.1613022221459283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,4096,0.01640266676743825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,3584,0.017130666308932833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,4096,0.014118222726715935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,3072,0.015527110960748462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,2048,0.010063110954231685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,2560,0.014168888330459595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,3584,0.013929777675204806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,3072,0.011976888610257043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,6144,0.017528888252046373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,3072,0.013806222213639153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,2560,0.012114666402339935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,1536,0.008768888811270395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,2048,0.011048888994587792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,5120,0.015525332755512662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,1024,0.006335999816656113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,1536,0.009033777647548253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,51200,0.10696622398164536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,768,0.00462666650613149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,1024,0.007678222325113084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,2048,0.010741333166758219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,512,0.0035342222286595237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,768,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,2048,0.009343999955389235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,256,0.002944888960984018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,512,0.007078222102589077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,512,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,256,0.006664000037643645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,1536,0.009670221971140968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,256,0.005647111270162795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,64,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,1024,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,32,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,128,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,64,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,65536,0.17129511303371855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,32,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,51200,0.12898843818240696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,65536,0.0887991123729282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,65536,0.07215733660591973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,16384,0.04482577906714546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,51200,0.0569102234310574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,3584,0.013051555388503604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,12288,0.033173332611719765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,16384,0.022501332892311945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,2560,0.010103111465771994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,10240,0.027830223242441814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,12288,0.027598222096761067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,10240,0.01664977769056956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,8192,0.023413333627912734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,16384,0.03836888737148709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,12288,0.018630221486091614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,8192,0.015404444601800708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,7168,0.02055022286044227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,10240,0.02399377855989668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,6144,0.01832444469134013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,7168,0.018068444397714403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,51200,0.06866133213043213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,5120,0.015777778294351365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,768,0.007367110914654202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,8192,0.02068088948726654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,5120,0.014825777875052558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,4096,0.01351199961370892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,128,0.0029760001020299066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,3584,0.012079111403889127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,128,0.006762666834725275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,6144,0.016528000434239704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,3072,0.011010666688283285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,3584,0.009738666315873465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,3072,0.008956444760163626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,2560,0.009679999616410997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,7168,0.01368622233470281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,4096,0.013450666434235044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,2560,0.010077333284748925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,2048,0.008346666892369589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,2560,0.00849333322710461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,6144,0.013092444174819522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,1536,0.007293333609898885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,2048,0.00795111142926746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,2048,0.009383999639087254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,1024,0.005401777724424998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,5120,0.011791110866599612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,3072,0.011419555379284753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,768,0.0038924444880750445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,1024,0.0070186663005087115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,1024,0.006720888945791457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,512,0.003567111161020067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,768,0.00674222202764617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,768,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,256,0.0032035555276605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,512,0.0064479998416370815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,3584,0.011745777395036487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,128,0.0028808888875775864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,256,0.006191111273235745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,256,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,1536,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,64,0.0028799999919202593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,128,0.0063368889192740125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,128,0.004951111144489712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,64,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,32,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,65536,0.08323199881447686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,65536,0.0661991106139289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,51200,0.117122663391961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,1536,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,32,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,51200,0.06277066469192505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,12288,0.03281777766015794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,16384,0.022164444128672283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,16384,0.038422223594453596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,10240,0.02840533190303379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,12288,0.030455999904208716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,512,0.00564177789621883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,8192,0.024311999479929607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,10240,0.02293066680431366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,16384,0.040919111834632024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,8192,0.020311110549502902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,51200,0.05412444472312927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,65536,0.1481893327501085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,6144,0.016864000095261466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,12288,0.017942221628295053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,6144,0.012130666938092975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,5120,0.014466666513019137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,10240,0.01590222285853492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,8192,0.014455111490355598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,5120,0.013717333475748697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,4096,0.010553778045707278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,7168,0.013064000341627332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,7168,0.018204443984561496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,3584,0.011169777976142036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,3584,0.009031111167536842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,6144,0.016190222567982145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,3072,0.01000088867213991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,3072,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,5120,0.01107377807299296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,2560,0.009002666506502364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,2560,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,2048,0.007675555845101674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,2560,0.009749333063761393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,4096,0.01279022213485506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,1536,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,2048,0.007416889071464539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,2048,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,1024,0.004559999952713649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,1536,0.007017778025733099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,1536,0.007344000041484833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,768,0.00388711111413108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,1024,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,7168,0.01882488860024346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,512,0.003527111063400904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,768,0.006329777754015393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,768,0.005987555616431766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,256,0.0029075555503368378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,512,0.006119110931952794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,4096,0.009708444277445475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,512,0.005629333357016246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,128,0.002867555452717675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,256,0.005984888722499211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,64,0.0028728888266616394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,128,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,32,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,128,0.004899555610285865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,65536,0.12223200003306071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,64,0.00526577772365676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,1024,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,32,0.0052888890107472735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,51200,0.0969360007180108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,65536,0.0635146697362264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,16384,0.035784890254338585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,51200,0.056108441617753774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,51200,0.05159644285837809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,12288,0.028622223271263972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,16384,0.021019554800457422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,3584,0.011164444188276926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,12288,0.017442666822009616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,12288,0.024720888998773363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,10240,0.02427733275625441
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,10240,0.0155404441886478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,65536,0.07195288605160184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,8192,0.02144177754720052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,8192,0.013824889229403602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,256,0.005223999834722943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,16384,0.029637333419587877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,7168,0.016531555189026725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,6144,0.017039111918873258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,10240,0.021035555336210463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,6144,0.0148044443792767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,5120,0.015254222684436373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,6144,0.01202666676706738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,7168,0.019038221902317472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,4096,0.011746666497654386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,5120,0.010483555495738983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,5120,0.013131555583741931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,3584,0.010419555836253695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,4096,0.009378666679064432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,4096,0.012343111137549082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,3072,0.009512888888518015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,3584,0.01072533345884747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,3072,0.00795288880666097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,3072,0.011083555718262991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,8192,0.018897778458065458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,2560,0.007686222592989604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,2560,0.009145778086450364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,7168,0.012887111140622033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,2048,0.008099555969238281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,3584,0.008271999657154083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,1536,0.00665511108107037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,1536,0.007348444726732042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,2560,0.00830577810605367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,1024,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,1024,0.006545777950021956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,768,0.003540444291300244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,768,0.006304889089531369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,512,0.0032071111102898917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,4096,0.012503110700183444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,512,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,256,0.003195555466744635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,1024,0.004576000074545543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,256,0.0057857777509424425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,2048,0.007285333342022366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,128,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,256,0.005239999956554837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,2048,0.0069884442620807225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,128,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,3072,0.011046222514576383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,64,0.005155555489990446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,512,0.005629333357016246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,32,0.0052471111218134565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,65536,0.10591644710964626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,51200,0.0842168860965305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,65536,0.06097866429222954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,64,0.0028497777465316984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,51200,0.04945333467589485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,16384,0.03167377909024557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,16384,0.02029511166943444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,16384,0.027523555689387854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,12288,0.022657778528001573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,12288,0.016861332787407767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,10240,0.021759111020300124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,768,0.00601244428091579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,10240,0.015142222245534262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,10240,0.01960266629854838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,8192,0.017885333961910672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,8192,0.013288000391589271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,1536,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,32,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,7168,0.012427555190192329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,8192,0.0196142229768965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,6144,0.01406044430202908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,6144,0.011409777734014722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,5120,0.014081777797804939
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,5120,0.012773333324326409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,5120,0.00981955561372969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,4096,0.011058666639857821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,65536,0.06634577777650622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,4096,0.008722667064931657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,3584,0.009518221848540837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,3584,0.010323555933104621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,12288,0.025468443830808003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,3072,0.008902221918106079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,3584,0.008351999852392409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,3072,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,3072,0.010212444596820408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,7168,0.017302221722073026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,2560,0.008008889026112026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,2048,0.006992888947327931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,6144,0.015765332513385348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,2048,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,1536,0.005290666802061929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,2560,0.007985778152942657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,51200,0.05084622237417433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,1024,0.003935111065705617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,2560,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,7168,0.015843555331230164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,1024,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,768,0.0035537779331207275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,768,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,768,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,512,0.003198222153716617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,1024,0.006600889066855113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,512,0.005983110931184556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,256,0.0029120000286234748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,1536,0.006884444504976273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,256,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,256,0.0053004444473319584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,128,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,128,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,64,0.0025315555847353404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,512,0.005617777920431561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,32,0.0028479999552170434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,64,0.005115555392371284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,32,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,65536,0.05881333351135254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,51200,0.06664444340599908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,2048,0.006994666324721442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,51200,0.04626577761438158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,51200,0.04758488800790575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,16384,0.0262080000506507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,1536,0.006981333096822103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,12288,0.020969778299331665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,16384,0.019659555620617337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,16384,0.02714755634466807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,10240,0.0181688881582684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,12288,0.022188444932301838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,10240,0.014455111490355598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,128,0.004977777600288391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,8192,0.012772444221708508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,8192,0.01754488878779941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,4096,0.01184977756606208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,7168,0.011728000309732227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,7168,0.015628443823920358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,6144,0.013481777575280933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,12288,0.016184000505341422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,5120,0.012067555553383298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,6144,0.010748444332016839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,8192,0.015878222054905362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,6144,0.014003554979960123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,5120,0.009176000124878353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,5120,0.011807999677128263
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,3584,0.011037333144081963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,4096,0.008362666600280339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,4096,0.01108000013563368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,65536,0.08989689085218643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,3584,0.008014221986134848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,3584,0.008968888885445064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,2560,0.008998221821255153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,65536,0.05859022008048164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,3072,0.00868088917599784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,2048,0.007631999750932057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,2560,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,2560,0.007667555577225155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,1536,0.005984888722499211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,2048,0.007661333514584436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,7168,0.01458666721979777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,1536,0.006642666541867786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,1024,0.005364444520738389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,1024,0.006495999793211619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,3072,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,768,0.004268444246715969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,768,0.006385777973466449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,512,0.003593777616818746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,10240,0.01843644513024224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,2048,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,256,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,512,0.0059262220230367445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,512,0.005624889085690181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,1536,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,256,0.0052871112194326185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,128,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,64,0.0028391112056043413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,768,0.00564533347884814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,32,0.0028888889484935333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,128,0.0029191111938820947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,65536,0.07795733213424683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,1024,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,128,0.004931555440028508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,51200,0.06084444125493368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,65536,0.05799466371536255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,4096,0.010784889260927836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,64,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,16384,0.024122666981485155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,51200,0.04680000080002678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,51200,0.04449866546524895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,12288,0.018914666440751817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,16384,0.019582221905390423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,10240,0.01643822259373135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,12288,0.016162667009565566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,12288,0.020276443825827706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,8192,0.014627556006113688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,10240,0.016501333978441026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,65536,0.05695822503831652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,8192,0.016120889120631747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,7168,0.013468444347381592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,3072,0.01036444471942054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,7168,0.011055111057228513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,10240,0.014464888307783337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,8192,0.012441777520709567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,6144,0.01260088880856832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,5120,0.011391999820868174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,6144,0.01036266651418474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,5120,0.00923999978436364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,4096,0.010125333236323463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,6144,0.012773333324326409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,4096,0.009534222384293875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,3584,0.010784889260927836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,4096,0.008174222376611497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,3584,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,3072,0.009379555781682333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,16384,0.028679112593332928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,3072,0.008392000363932716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,3072,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,2560,0.007461333440409766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,2048,0.007299555672539606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,2560,0.007624000310897827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,3584,0.007793777518802219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,1536,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,256,0.00562755556570159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,2560,0.008671999805503422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,1536,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,1024,0.005635555419656966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,1024,0.006299555715587404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,768,0.004243555582231945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,32,0.004960000101062986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,768,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,1536,0.006375999914275275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,512,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,512,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,2048,0.0069395556218094295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,7168,0.014526221487257214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,256,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,1024,0.004558222161398994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,128,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,128,0.005664888769388199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,256,0.003076444483465619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,5120,0.011417778001891242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,32,0.002696000039577484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,256,0.005061333378156026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,128,0.0026382222357723448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,65536,0.062346663739946157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,32,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,51200,0.050208889775806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,65536,0.05696800020005968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,65536,0.053957333167394005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,16384,0.02028888960679372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,51200,0.045907557010650635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,64,0.002566222308410539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,12288,0.017190222938855488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,16384,0.01924622224436866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,768,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,12288,0.015827554795477126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,10240,0.01499644418557485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,10240,0.014124444789356656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,8192,0.01312444441848331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,51200,0.04267999860975477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,8192,0.01514933341079288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,2048,0.007329777710967594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,7168,0.01239999963177575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,16384,0.022493332624435425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,8192,0.011491555306646558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,6144,0.011383111278216044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,64,0.004967111266321606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,5120,0.01053066634469562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,6144,0.010260444548394945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,6144,0.011782222323947482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,5120,0.00907466643386417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,10240,0.015489778584904142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,5120,0.010384000009960597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,3584,0.009008000294367472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,4096,0.00831200016869439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,4096,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,3072,0.008823110825485652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,3584,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,3584,0.007704888780911763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,2560,0.007667555577225155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,3072,0.007636444436179266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,7168,0.010220444036854638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,3072,0.007694222033023834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,2048,0.00630044440428416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,2560,0.007348444726732042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,1536,0.005909333212508096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,2560,0.006643555644485686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,2048,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,1024,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,1536,0.006675555474228329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,1536,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,768,0.004260444392760594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,1024,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,1024,0.005648889061477449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,768,0.006275555739800136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,768,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,512,0.0039013334446483185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,12288,0.01958311100800832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,7168,0.013154666456911298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,256,0.0034942221310403612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,512,0.005292444593376583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,2048,0.006981333096822103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,4096,0.009688888986905416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,128,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,64,0.002518222149875429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,256,0.005299555758635203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,32,0.002946666752298673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,512,0.0038942222793896994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,64,0.004943999979231092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,65536,0.05698933204015096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,512,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,32,0.005045333256324132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,65536,0.05173066589567396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,51200,0.04518666532304552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,128,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,16384,0.018838221828142803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,51200,0.041449778609805636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,16384,0.018907555275493197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,12288,0.016210666961140104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,65536,0.05609777900907728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,256,0.005660444498062134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,10240,0.013722666435771518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,10240,0.01309511148267322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,8192,0.012138666378127204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,8192,0.01350488927629259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,8192,0.011411555939250521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,7168,0.011409777734014722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,7168,0.010405333505736457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,6144,0.01071733319097095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,10240,0.01514311134815216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,6144,0.009667555491129557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,6144,0.010375111467308467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,5120,0.009716444545321995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,16384,0.021243555678261652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,4096,0.009337777892748514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,5120,0.009002666506502364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,5120,0.009080000221729279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,51200,0.04472000069088406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,4096,0.008268444074524773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,4096,0.008239111138714684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,3072,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,3584,0.007975111405054728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,3584,0.007303111255168915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,2560,0.0069431112044387394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,3072,0.007663999994595845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,2560,0.007086222370465596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,2048,0.005994666781690385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,128,0.0029031110720502008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,2048,0.006321777900060018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,2048,0.00700800038046307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,1536,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,3072,0.0075493330756823225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,1024,0.004571555389298333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,1536,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,1024,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,768,0.004265777766704559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,12288,0.01574488812022739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,768,0.005446222093370225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,768,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,7168,0.0123511114054256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,1536,0.006602666858169768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,256,0.003261333331465721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,512,0.005274666680230035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,2560,0.006955555743641323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,128,0.0031644445326593188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,12288,0.018925334016482036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,128,0.004899555610285865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,128,0.00564444437623024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,512,0.006047111004590988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,64,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,32,0.0028524444335036804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,65536,0.04966844452752007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,1024,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,32,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,3584,0.008788444101810455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,65536,0.050385776493284434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,512,0.003589333345492681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,16384,0.015801777442296345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,51200,0.039905776580174766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,16384,0.01833777791923947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,12288,0.01310400002532535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,256,0.005646222167544895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,51200,0.0381644434399075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,12288,0.014518222875065274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,10240,0.011916444533401065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,64,0.0026319999661710528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,10240,0.012712000144852532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,10240,0.01348977784315745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,8192,0.011170667078759937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,8192,0.011955555942323474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,7168,0.010830221904648675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,256,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,7168,0.010444444086816577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,7168,0.010383110907342698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,65536,0.05549066596561008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,5120,0.008985777695973715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,6144,0.009703999592198266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,6144,0.008789333204428354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,51200,0.044887112246619336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,4096,0.010194666683673859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,5120,0.008700444466537898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,16384,0.02123822271823883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,4096,0.008299555215570662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,4096,0.007634667058785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,6144,0.009695111049546136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,3584,0.008038221961922115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,12288,0.015266666809717814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,3584,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,2560,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,3072,0.007742221984598372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,3072,0.0069439998931354946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,2048,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,2560,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,2048,0.006693333387374878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,3584,0.009027555584907532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,3072,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,1536,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,1024,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,8192,0.011376000112957425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,2560,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,768,0.004192000048028098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,1024,0.006507555643717448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,768,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,2048,0.0063279999627007385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,768,0.0053022222386466126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,512,0.0035386664999855887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,256,0.0032435556252797446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,512,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,512,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,256,0.005589333259397083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,256,0.004940444396601783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,64,0.002874666617976295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,128,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,1024,0.005621333503060871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,64,0.004956444518433677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,1536,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,65536,0.03621866636806064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,65536,0.05488711264398363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,65536,0.04847466614511278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,128,0.004964444372389051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,51200,0.044575111733542554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,16384,0.013160000244776407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,51200,0.02881511052449544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,16384,0.016753777861595154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,12288,0.011228444675604502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,5120,0.009165333376990424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,12288,0.01350488927629259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,12288,0.014654222461912366
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,10240,0.012436444560686747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,10240,0.012106666962305704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,8192,0.010313777459992303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,32,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,8192,0.011062222222487131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,7168,0.009690666364298927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,7168,0.009564444422721863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,1536,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,6144,0.008970666262838576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,7168,0.010377777947319878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,51200,0.03866666555404663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,5120,0.007826666865083907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,6144,0.008683555656009251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,10240,0.010440888504187265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,4096,0.007640889121426477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,5120,0.008207110895050896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,4096,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,16384,0.01779822177357144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,4096,0.007803555164072249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,6144,0.009679999616410997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,3072,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,3584,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,32,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,2560,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,128,0.0029688889367712867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,2560,0.006817777951558431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,2048,0.005967999911970562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,2560,0.007375111182530721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,3584,0.008103999826643202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,2048,0.006953777538405524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,2048,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,8192,0.00981955561372969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,1536,0.006653333289755716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,1024,0.004226666771703296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,1024,0.006304889089531369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,768,0.003906666818592283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,768,0.005494222044944763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,768,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,5120,0.00868800034125646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,512,0.003563555578390757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,3072,0.0069466663731469055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,256,0.0029528888149393927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,512,0.005963555640644497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,512,0.005063999858167437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,1024,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,256,0.005664000080691443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,256,0.004893333133724001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,64,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,128,0.005590222362014983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,128,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,1536,0.0052426668504873914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,32,0.002560888934466574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,65536,0.025581333372328017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,3072,0.007702222300900354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,1536,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,32,0.005025777965784073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,65536,0.04661422305636936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,65536,0.054915553993648954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,3584,0.007148444652557373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,16384,0.01462577780087789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,51200,0.04394222299257914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,12288,0.012079999678664737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,51200,0.03771911064783732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,16384,0.016000888413853116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,10240,0.011895111037625207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,12288,0.013606222139464485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,51200,0.021223111285103693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,8192,0.010645333263609143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,10240,0.011095999843544431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,128,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,7168,0.009509333305888707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,8192,0.0099973330895106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,6144,0.008581333690219456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,7168,0.010384000009960597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,7168,0.009377777576446533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,64,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,6144,0.008715555899673039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,10240,0.01221777747074763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,5120,0.008999110923873054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,5120,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,4096,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,8192,0.010828444527255164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,4096,0.007647111184067196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,16384,0.016211556063758004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,3072,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,3584,0.007336888876226213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,12288,0.012810666528013019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,2560,0.00563822231358952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,3072,0.007627555893527136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,3072,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,2048,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,2560,0.006599111275540457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,3584,0.006960888703664144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,2048,0.006130666782458623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,3584,0.008023111356629265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,6144,0.009711110757456886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,1536,0.005759999983840519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,1024,0.004231111043029361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,1024,0.006302222195598815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,2560,0.007325333025720384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,1024,0.005495111147562663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,768,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,512,0.0035457776652442086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,768,0.005361777626805835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,512,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,256,0.0031999999450312722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,256,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,256,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,128,0.0028693332440323303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,512,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,128,0.005617777920431561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,64,0.002663111107216941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,4096,0.0074106670088238185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,32,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,2048,0.006981333096822103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,1536,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,32,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,5120,0.007705777883529663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,768,51200,0.017903111047214933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,768,0.003948444293604957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,51200,0.03701600101259019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,768,16384,0.010372444159454769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,51200,0.043508443567487926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,768,65536,0.021340444684028625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,768,12288,0.011072888970375061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,16384,0.015899555550681222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,16384,0.014860444598727755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,768,10240,0.009471110999584198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,1536,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,12288,0.012096888489193387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,768,8192,0.009004444711738164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,10240,0.011058666639857821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,8192,0.010753778119881948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,768,7168,0.00833777752187517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,7168,0.00924711094962226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,128,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,768,6144,0.007652444144090016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,7168,0.010024000373151567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,10240,0.012089778151777057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,768,5120,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,6144,0.009370666411187913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,6144,0.00867377801073922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,768,4096,0.007296889192528195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,64,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,4096,0.008008889026112026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,768,3584,0.0069822221994400024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,12288,0.013457777599493662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,3584,0.007295110987292395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,3584,0.008016000191370646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,768,3072,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,8192,0.009824000298976898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,5120,0.008000888758235509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,768,2560,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,3072,0.007625777688291337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,4096,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,2560,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,2560,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,2048,0.006999111009968652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,2048,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,768,1536,0.0052373334765434265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,1536,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,5120,0.00871022211180793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,768,1024,0.0042204442951414315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,3072,0.006933333145247565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,768,768,0.0038124442928367188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,1024,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,1024,0.005568888866239124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,768,2048,0.0052471111218134565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,768,0.005977777971161737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,65536,0.05412888858053419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,768,512,0.0033866665843460294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,768,0.005240889059172736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,512,0.005909333212508096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,768,128,0.0028479999552170434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,256,0.005585777676767773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,768,64,0.0028782222006056043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,256,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,768,32,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,128,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,512,65536,0.01685333251953125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,64,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,32,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,512,51200,0.01516266663869222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,65536,0.045831110742357045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,65536,0.04551288816663954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,768,256,0.0032879999942249725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,51200,0.038712001509136625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,512,0.005229333208666907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,51200,0.03666222095489502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,512,12288,0.00868888861603207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,16384,0.015846222639083862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,128,0.005614222337802251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,512,10240,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,12288,0.013584888643688627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,512,8192,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,12288,0.012200888660218982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,10240,0.010782221953074137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,8192,0.010767111347781287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,65536,0.051593777206208974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,512,16384,0.011747555600272285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,7168,0.00907555553648207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,16384,0.014744000302420722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,6144,0.008845333423879411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,512,5120,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,10240,0.012094222009181976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,6144,0.00963644435008367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,512,4096,0.007348444726732042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,512,7168,0.007648888561460707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,8192,0.010074666804737514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,512,3584,0.006953777538405524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,4096,0.00813422186507119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,4096,0.007672888537247975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,512,6144,0.007292444507280986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,3584,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,3584,0.006959111326270633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,512,2560,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,3072,0.007371555599901412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,1536,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,3072,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,2560,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,2560,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,512,1536,0.004558222161398994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,2048,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,2048,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,512,1024,0.0038977778620190094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,1536,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,7168,0.010040889183680216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,512,3072,0.00666133314371109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,512,768,0.003547555456558863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,1024,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,512,512,0.003330666571855545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,768,0.005963555640644497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,768,0.00526311124364535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,512,2048,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,512,256,0.0029022221763928733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,512,0.005049777941571342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,512,128,0.0028799999919202593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,256,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,128,0.005662222289376789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,1536,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,1024,0.006295111030340195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,512,32,0.0025964443468385274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,64,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,256,65536,0.015805333852767944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,512,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,256,0.005676444619894028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,65536,0.046289778418011136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,256,51200,0.013150222599506378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,65536,0.04254577888382805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,51200,0.03748266564475165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,512,64,0.0025742221623659134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,256,16384,0.009071110851234859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,16384,0.01585422290696038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,256,12288,0.008375110725561777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,32,0.00489688871635331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,5120,0.008963555925422246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,12288,0.012402666939629449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,5120,0.008352888955010308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,10240,0.012082666986518435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,128,0.004910222358173794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,256,8192,0.007000889215204451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,8192,0.010737777584128909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,8192,0.009687999884287516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,256,7168,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,7168,0.010086221827401055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,7168,0.009356444080670675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,256,6144,0.006335111127959357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,6144,0.009576888548003303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,256,5120,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,51200,0.03406755460633172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,5120,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,256,4096,0.005699555493063397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,16384,0.014481777946154276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,4096,0.008344000412358178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,256,3584,0.0063271112740039825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,12288,0.01311822235584259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,256,10240,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,3584,0.007740444607204861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,3584,0.007168889045715332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,256,2560,0.006292444550328785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,10240,0.01108888867828581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,2560,0.007353777686754863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,5120,0.008295999632941352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,4096,0.0069857777820693115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,256,3072,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,256,1536,0.004224888980388641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,2048,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,6144,0.008674666285514832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,1536,0.006309333360857434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,1536,0.005591111050711737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,2560,0.006643555644485686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,1024,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,2048,0.00608800020482805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,768,0.0052871112194326185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,256,512,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,768,0.0062977779242727495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,3072,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,3072,0.006712889091836081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,512,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,256,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,256,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,128,0.004910222358173794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,128,0.005591111050711737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,256,64,0.002518222149875429
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,512,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,256,32,0.0025617778301239014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,64,0.0048942222363419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,256,128,0.0028488888508743713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,32,0.004945777770545748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,65536,0.045999109745025635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,65536,0.04179822074042427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,256,256,0.0028968888024489083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,51200,0.0336079994837443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,128,16384,0.007700444095664554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,51200,0.037372443411085345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,1024,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,256,768,0.003555555724435382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,16384,0.015895111693276297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,16384,0.01423555612564087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,128,10240,0.006981333096822103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,12288,0.01314933349688848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,12288,0.012117333710193634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,128,51200,0.012438221938080259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,128,8192,0.0069475554757648045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,10240,0.012094222009181976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,128,7168,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,10240,0.010779555473062726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,256,1024,0.0038319999972979226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,128,6144,0.006561777657932705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,7168,0.009034666750166152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,256,2048,0.0059262220230367445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,128,12288,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,6144,0.009407999614874521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,5120,0.008950221869680617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,5120,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,8192,0.01072266697883606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,4096,0.007096889118353526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,128,3584,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,4096,0.008284444610277811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,3584,0.00797155582242542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,128,3072,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,128,4096,0.0058853332367208265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,3072,0.006679111056857639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,128,2560,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,3072,0.0073724447025193115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,8192,0.009728888670603434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,2560,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,128,2048,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,2560,0.00664533343580034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,2048,0.006908444480763541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,128,65536,0.014118222726715935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,128,1536,0.00423911131090588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,2048,0.006162666612201267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,128,1024,0.003912000192536248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,1536,0.006329777754015393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,1536,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,128,768,0.003363555504216088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,1024,0.005643555687533484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,768,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,7168,0.010052444206343757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,128,5120,0.00601244428091579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,3584,0.0069955554273393415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,128,256,0.0028488888508743713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,1024,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,256,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,128,128,0.0028568889117903183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,768,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,128,512,0.0032337777730491427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,128,64,0.00254844439526399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,128,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,128,32,0.0028080000645584534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,6144,0.008346666892369589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,64,65536,0.010738666686746808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,32,0.004933333231343163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,64,51200,0.009735999835862054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,65536,0.04167822334501478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,256,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,64,12288,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,128,0.004683555414279302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,16384,0.014147554834683737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,12288,0.012099555797047086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,64,8192,0.006607111129495833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,512,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,64,7168,0.005985777825117111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,10240,0.010780444575680627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,64,6144,0.0064560001095136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,8192,0.009690666364298927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,64,5120,0.005922666854328579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,7168,0.009168888959619734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,64,4096,0.00563733321097162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,51200,0.03359555535846286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,64,3584,0.006238222122192383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,5120,0.007991111112965478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,64,10240,0.0059262220230367445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,4096,0.0075502221783002215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,64,2560,0.005977777971161737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,64,0.00507733349998792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,64,2048,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,2560,0.006663110935025745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,64,1536,0.004263111286693149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,2048,0.00612088872326745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,64,1024,0.0038933331767717996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,1536,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,512,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,1024,0.005647111270162795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,64,512,0.0032195556494924757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,3072,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,64,256,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,512,0.005306666509972678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,64,128,0.0028799999919202593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,6144,0.008685333033402761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,64,64,0.0026808888134029177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,64,3072,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,128,0.004947555561860402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,64,32,0.002692444456948174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,3584,0.0069644442862934535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,32,0.004968889057636261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,768,0.005493333356248007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,32,51200,0.008708444734414419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,65536,0.041781332757737905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,51200,0.033378667301601835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,256,0.004805333498451444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,32,12288,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,32,10240,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,64,768,0.0035573335157500375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,32,65536,0.009352000223265754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,10240,0.0107022225856781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,32,7168,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,8192,0.009699555734793345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,32,16384,0.006647999915811751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,32,6144,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,32,5120,0.006256888724035687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,7168,0.008719999757077958
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,32,4096,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,16384,0.014460444450378418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,32,3584,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,5120,0.007866666548781926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,32,3072,0.005639111002286275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,4096,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,32,8192,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,32,2560,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,3584,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,3072,0.006640000061856375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,32,1536,0.004223111189073986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,64,16384,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,2048,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,1536,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,32,768,0.0032720000793536506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,12288,0.012087110843923358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,32,2048,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,32,512,0.00319377767542998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,2560,0.006664888726340399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,32,256,0.0028924445311228433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,512,0.00508088908261723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,32,128,0.0025759999536805684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,6144,0.008656000097592672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,32,64,0.0028799999919202593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,128,0.004590222405062782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,64,0.00489777781897121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,32,32,0.0025128889828920364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,768,0.005292444593376583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,32,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,5,32,1024,0.0038782221575578055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,51200,0.4859413570827908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,51200,1.0126355489095051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,64,0.004952888935804367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,16384,0.25591911209954155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,51200,0.4452826711866591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,12288,0.2564302285512288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,16384,0.3399804433186849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,16384,0.1366951068242391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,10240,0.21668710973527697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,12288,0.10258577929602729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,16384,0.16941866609785292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,1024,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,10240,0.1315173308054606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,8192,0.1695528825124105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,10240,0.08585333161883885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,12288,0.15247821807861328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,8192,0.11798489093780518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,7168,0.14709066020117864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,8192,0.0700204438633389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,7168,0.10228888856040107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,10240,0.10796711179945205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,256,0.004978666702906291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,7168,0.06206399864620633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,6144,0.127691560321384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,5120,0.10239199797312419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,6144,0.05456711186303032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,7168,0.07913955714967516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,12288,0.12616356213887533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,6144,0.1033395528793335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,4096,0.08611733383602566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,5120,0.08016444577111138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,6144,0.0690808892250061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,4096,0.06876444154315524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,4096,0.04050044549836053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,5120,0.059531556235419385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,4096,0.04959555466969808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,3072,0.06342133548524645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,3584,0.03504088852140639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,3584,0.045203556617101036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,3072,0.05137599839104546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,5120,0.046895109944873385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,3072,0.03987555702527364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,3584,0.07273955477608575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,2560,0.04048000110520257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,2048,0.04347999890645345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,2560,0.02880000074704488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,51200,0.49436357286241317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,2048,0.03666222095489502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,1536,0.03386755453215705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,2560,0.05349599983957079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,3072,0.0314408904976315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,1536,0.02385422256257799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,1536,0.02780533168050978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,1024,0.02411288850837284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,1536,0.026250667042202417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,2560,0.034938666555616595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,768,0.018630221486091614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,3584,0.0539608895778656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,1024,0.020308444897333782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,1024,0.020510221521059673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,512,0.013728888498412238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,768,0.016920000314712524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,2048,0.02829422222243415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,768,0.01753866672515869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,256,0.008472888833946651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,512,0.013111111190583972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,512,0.015127999915017022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,8192,0.0905039972729153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,128,0.006436444405052397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,256,0.010675555302037133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,256,0.01072088877360026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,256,0.012500444220172035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,1024,0.018228444788191054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,64,0.005587555468082428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,128,0.010607110957304636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,32,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,768,0.015537778536478678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,64,0.009030222064918941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,2048,0.03126310971048143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,512,0.012881778180599213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,65536,0.5130053096347386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,65536,0.971443600124783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,128,0.008201777935028076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,51200,0.7576577928331164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,32,0.00924177798959944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,51200,0.38831289609273273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,16384,0.25126756562127006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,65536,0.42278133498297793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,16384,0.14569066630469427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,12288,0.19226845105489096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,51200,0.3372942341698541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,128,0.010423111418883005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,16384,0.11094844341278076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,10240,0.16104711426628962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,12288,0.1255884435441759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,12288,0.0852782196468777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,10240,0.12381688753763835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,10240,0.07226577732298109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,8192,0.13253955046335855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,10240,0.08724977572758992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,8192,0.056392000781165234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,8192,0.07313866747750176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,7168,0.11619822184244792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,16384,0.13573511441548666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,12288,0.101310220029619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,6144,0.10123733017179702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,7168,0.05863022141986423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,8192,0.1147528886795044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,6144,0.08181422286563449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,7168,0.06434666448169284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,6144,0.04591822293069628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,6144,0.0566799971792433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,4096,0.07270399729410808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,51200,0.3872888882954915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,65536,0.5106924374898275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,7168,0.08335466517342462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,5120,0.04921066761016846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,3584,0.0641875531938341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,4096,0.055721779664357506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,4096,0.033999999364217125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,4096,0.041407111618253924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,3584,0.029658665259679157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,3584,0.03862133291032579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,3072,0.04230044285456339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,2560,0.0467217763264974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,5120,0.0649075574345059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,3072,0.03438399897681342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,5120,0.03971555497911241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,2048,0.03825511203871833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,2560,0.03153689040078057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,3072,0.05635999970965915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,5120,0.09028266535864936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,1536,0.029601779248979356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,2048,0.027200000153647527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,3072,0.026126222478018865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,1536,0.02306933369901445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,1024,0.02161955502298143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,1536,0.024128889044125874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,2560,0.024167999625205994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,1024,0.01577333278126187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,768,0.015848888291252982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,3584,0.04384088847372267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,2048,0.02353333267900679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,768,0.014607111612955729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,768,0.013790222505728403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,768,0.015862221519152325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,1536,0.020435555113686454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,2560,0.0359919998380873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,256,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,512,0.013163555827405719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,1024,0.019464888506465487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,256,0.008604444563388824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,2048,0.03014311194419861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,128,0.004229333251714706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,256,0.011339555184046427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,512,0.011359111302428775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,64,0.0035422220826148987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,128,0.007322666545708974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,32,0.003589333345492681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,128,0.010408889088365767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,512,0.011413333316644033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,64,0.008282666405042013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,1024,0.017154667112562392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,65536,0.32156533665127224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,65536,0.17540710502200654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,51200,0.24951110945807564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,65536,0.1524462170071072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,128,0.009680888719028896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,51200,0.13024532794952393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,512,0.01146399974822998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,51200,0.12272000312805176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,32,0.008336000144481659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,16384,0.0656906697485182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,12288,0.06633333365122478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,16384,0.04229066769282023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,256,0.009677333136399588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,16384,0.052490668164359204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,12288,0.033188445700539485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,65536,0.17043377293480766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,12288,0.04010044535001119
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,16384,0.08634399705462986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,10240,0.04538844360245598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,51200,0.13859200477600098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,10240,0.036370668146345354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,7168,0.03761777612898085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,8192,0.023856888214747112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,8192,0.036772443188561335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,8192,0.032101333141326904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,7168,0.033372445238961115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,10240,0.05774133072959053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,7168,0.02774133284886678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,12288,0.05088177654478285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,6144,0.03186310993300544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,5120,0.027956446011861164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,6144,0.025212445192866858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,5120,0.026118222210142348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,5120,0.022727999422285292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,4096,0.024293333292007446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,4096,0.022239110536045496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,6144,0.02033333314789666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,4096,0.01552177800072564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,4096,0.019702222612169053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,5120,0.017705778280893963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,3584,0.01887911061445872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,3584,0.014490667316648694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,10240,0.02888800038231744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,3584,0.018479110466109384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,3072,0.013170666992664337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,3072,0.01779911087618934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,3584,0.020630222227838304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,3072,0.017131555411550734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,7168,0.022180444664425317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,2560,0.014480888843536377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,2560,0.015787555111779105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,8192,0.042733334832721286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,3072,0.017906667457686532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,2048,0.013620444469981723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,1536,0.010709332923094431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,2048,0.01420088940196567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,2048,0.013089777694808112
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,1024,0.008403555386596257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,2560,0.015624889069133334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,1536,0.011387555963463254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,1536,0.011431111229790581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,768,0.007116444408893585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,1024,0.009287999735938178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,6144,0.03248622351222568
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,2048,0.01333688861793942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,1024,0.011701333026091257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,512,0.0047306666771570844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,768,0.007696888513035244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,768,0.010036444498433007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,512,0.007432888779375289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,512,0.0069848886794514125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,512,0.009001777403884465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,256,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,128,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,256,0.009024000002278222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,2560,0.012747555143303342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,128,0.005281777845488654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,128,0.006955555743641323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,128,0.008758222063382467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,32,0.0028773333049482773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,256,0.003552888830502828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,64,0.00591822216908137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,65536,0.2431795597076416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,256,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,1536,0.013087111214796701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,65536,0.1427439981036716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,1024,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,51200,0.18913066387176514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,65536,0.12962578402625188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,16384,0.06622222397062513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,51200,0.11167466640472412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,51200,0.1059066653251648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,32,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,16384,0.037321776151657104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,16384,0.04131999942991469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,65536,0.13358755906422934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,51200,0.10201778014500935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,12288,0.043600890371534556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,768,0.00833422193924586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,12288,0.03245866629812453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,10240,0.03988177908791436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,8192,0.038669334517584905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,16384,0.05630044142405192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,12288,0.051473776499430336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,8192,0.03193688723776076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,8192,0.021820444199774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,8192,0.02571288910177019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,10240,0.04447199900945028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,64,0.0028764444092909494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,6144,0.030200001266267564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,7168,0.020289777053727042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,7168,0.02304711110062069
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,10240,0.029025776518715754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,5120,0.026518222358491685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,6144,0.021078222327762183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,7168,0.03449689017401801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,5120,0.01899377836121453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,5120,0.016210666961140104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,4096,0.02271822260485755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,7168,0.028669334120220605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,10240,0.02572800053490533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,4096,0.014186667071448432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,4096,0.019854222734769184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,4096,0.017218665944205392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,3584,0.01690933273898231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,5120,0.02311377724011739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,3584,0.013167111410035027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,12288,0.029723554849624634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,3072,0.01663644446267022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,3584,0.020432000358899433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,3072,0.014936889211336771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,2560,0.015480889214409722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,6144,0.01824977828396691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,2560,0.011570666399266986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,2560,0.013444444371594323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,2048,0.0141004439857271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,3584,0.016223111086421542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,2048,0.01277955538696713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,2048,0.011437333292431302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,1536,0.008916444248623317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,6144,0.025644444757037695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,3072,0.012375999655988483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,1024,0.0070293330483966405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,1024,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,2048,0.013819555441538492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,768,0.006239999913507038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,1024,0.008371555142932469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,1536,0.011014222270912595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,1536,0.010049777726332346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,768,0.007676444119877285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,768,0.007985778152942657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,768,0.009396444592210982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,3072,0.018132444885041978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,512,0.006719111154476802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,512,0.00868711123863856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,1024,0.010103999740547603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,256,0.008391111261314815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,128,0.002903999967707528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,512,0.0038924444880750445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,128,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,512,0.007335999773608313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,256,0.0032115555885765287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,64,0.0029004443850782183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,32,0.0028702221396896574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,256,0.005975111077229182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,32,0.005971555494599872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,256,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,1536,0.012079999678664737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,65536,0.14194311036003962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,2560,0.014463111758232117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,65536,0.12781066364712185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,128,0.008358221915033128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,128,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,64,0.005617777920431561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,51200,0.10655111074447632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,65536,0.2040711111492581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,51200,0.10313600301742554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,12288,0.04424622323777941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,16384,0.03728888763321771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,12288,0.03579466541608175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,12288,0.029273778200149536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,16384,0.05793066819508871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,10240,0.038968000147077776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,65536,0.1123831139670478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,10240,0.031874666611353554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,8192,0.03289511137538486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,10240,0.02460711035463545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,51200,0.1601439979341295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,8192,0.029311110575993855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,7168,0.02960088849067688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,8192,0.024728000164031982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,7168,0.02436888880199856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,16384,0.04739288820160759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,10240,0.02641955514748891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,16384,0.03634044528007507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,6144,0.02201955517133077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,8192,0.02346044447686937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,6144,0.0177484436167611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,5120,0.018684445155991446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,6144,0.026404443714353774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,51200,0.08991733524534438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,4096,0.020033778415785897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,5120,0.015511110424995422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,5120,0.0177208897140291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,12288,0.027765333652496338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,4096,0.014190221826235453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,4096,0.016013332539134555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,3584,0.014135110709402295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,7168,0.021057777934604224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,7168,0.019444444113307528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,3584,0.013108444710572561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,3072,0.013659555878904132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,3072,0.01458044515715705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,3584,0.017836444907718234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,2560,0.012093333734406365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,3072,0.015875554747051664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,2560,0.011140444212489657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,2560,0.013391111459996967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,6144,0.019698666201697457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,5120,0.022825777530670166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,2048,0.011056000159846412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,1536,0.009230222139093611
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,2048,0.010036444498433007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,2048,0.013092444174819522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,3584,0.01553244392077128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,1536,0.00902133352226681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,2048,0.010371555884679159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,4096,0.01651822196112739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,3072,0.011680888632933298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,1024,0.007694222033023834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,768,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,1024,0.009707555174827576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,1536,0.009374221993817223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,768,0.007024889190991719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,512,0.0035742223262786865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,768,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,1536,0.012135110795497894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,768,0.009006222089131674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,256,0.0029306666304667792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,512,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,2560,0.013888888888888888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,1024,0.00814400033818351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,256,0.00563822231358952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,256,0.006676444576846228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,128,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,128,0.008990222381220924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,1024,0.00693866651919153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,64,0.0025804444319672054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,32,0.0028488888508743713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,512,0.00699733363257514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,512,0.0086666668454806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,65536,0.16922577222188315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,32,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,256,0.009002666506502364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,51200,0.12848622269100612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,65536,0.07197866837183635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,51200,0.06850222084257337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,51200,0.05829599830839369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,65536,0.10048533148235744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,64,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,128,0.0028515555378463534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,16384,0.022828444838523865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,12288,0.03385244475470649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,16384,0.03311999970012241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,65536,0.08999911281797622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,10240,0.029422223567962646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,12288,0.028078221612506445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,12288,0.027388445205158655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,16384,0.04674666788842943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,8192,0.024522667129834492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,10240,0.02475555572244856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,10240,0.024546666277779475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,51200,0.08043022288216485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,8192,0.021328000558747187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,8192,0.014807111687130399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,12288,0.018826666805479262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,7168,0.01903377804491255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,128,0.006340444501903322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,7168,0.02028622229894002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,7168,0.014499555031458536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,10240,0.017041777571042378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,6144,0.014262222581439547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,5120,0.015855999456511605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,16384,0.03470755616823832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,7168,0.021057777934604224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,8192,0.02233244478702545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,5120,0.014819555812411837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,5120,0.01718844473361969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,4096,0.014088888963063559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,3584,0.012435555458068848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,3584,0.012255999777052136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,4096,0.013433777623706393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,5120,0.011735111474990845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,3072,0.011008888483047485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,3584,0.014911111858155994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,6144,0.018499554859267343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,4096,0.015341333217091031
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,3072,0.011653333074516721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,2560,0.009357333183288574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,3072,0.009030222064918941
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,6144,0.01863822175396813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,2560,0.010377777947319878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,2048,0.008658666577604082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,2560,0.012947555217478009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,2048,0.008273777862389883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,2048,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,1536,0.007062222394678328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,2048,0.012693333129088083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,4096,0.011075555450386472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,1024,0.0058248887459437055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,1536,0.007891555627187094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,3072,0.014208000567224292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,2560,0.008780444661776224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,768,0.003921777837806278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,1024,0.009357333183288574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,3584,0.009736888938479954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,768,0.00923911150958803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,768,0.006616000086069107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,512,0.0035413333939181436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,1536,0.007330666813585494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,1536,0.010608889162540436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,256,0.0032151111712058387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,512,0.0059057776298787855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,6144,0.016807110773192514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,256,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,256,0.008014221986134848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,256,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,128,0.0031555555760860443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,128,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,128,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,64,0.002622222320901023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,1024,0.00703999979628457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,32,0.0028906667398081887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,64,0.0052675555149714155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,32,0.0053031109273433685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,65536,0.14770221710205078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,65536,0.08123999834060669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,128,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,512,0.008465777668688033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,51200,0.11714755164252387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,65536,0.06598755386140612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,51200,0.06223999791675144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,16384,0.04272888766394722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,51200,0.07966222365697224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,16384,0.033328887489106916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,12288,0.03416088885731167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,16384,0.021820444199774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,12288,0.027082666754722595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,768,0.006330666856633292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,10240,0.02901600135697259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,12288,0.01789066692193349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,10240,0.023400889502631292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,512,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,10240,0.02443733314673106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,16384,0.032933334509531655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,8192,0.020284444093704224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,65536,0.09910400046242608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,8192,0.020993777447276648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,7168,0.019095111224386428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,7168,0.019311999281247456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,8192,0.024501333634058636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,6144,0.016958221793174744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,7168,0.013295999831623502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,6144,0.016539555456903245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,12288,0.026745777991082933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,51200,0.05389155613051521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,6144,0.012080888781282636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,5120,0.014821334017647637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,7168,0.018447111050287884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,5120,0.01720355616675483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,1024,0.006698666761318843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,4096,0.012764444781674279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,4096,0.015370666980743408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,4096,0.009684444301658207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,3584,0.011735999749766456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,4096,0.012431110772821637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,10240,0.015957333975368075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,3072,0.010060444474220276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,3584,0.009025777379671732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,3584,0.014827556080288358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,3072,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,3072,0.011086222198274402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,2560,0.009016888837019602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,3072,0.013762666947311826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,2560,0.008035555481910706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,2560,0.010070222119490305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,2560,0.013384888569513956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,5120,0.011400000088744693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,2048,0.008761777646011775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,1536,0.006758222149478064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,2048,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,5120,0.0148044443792767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,1536,0.007641777396202087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,1024,0.004573333180612988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,1536,0.006933333145247565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,2048,0.007791111038790808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,1024,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,768,0.0038888889054457345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,1024,0.00902844468752543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,2048,0.012087110843923358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,768,0.006061333335108227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,768,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,8192,0.014140443669425117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,1536,0.010781333678298526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,512,0.005882666756709416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,512,0.0063288890653186375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,256,0.0028728888266616394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,512,0.008711111214425828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,1024,0.006656000183688269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,128,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,256,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,256,0.0052915554907586836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,256,0.00867733359336853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,64,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,128,0.00609688874748018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,128,0.004962666581074397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,32,0.0028906667398081887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,3584,0.011454222102959951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,65536,0.12463200092315674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,64,0.0052728888889153796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,32,0.005241777747869492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,51200,0.09856088956197102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,65536,0.06331111325158013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,51200,0.056277334690093994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,65536,0.09870222542021011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,512,0.0035413333939181436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,128,0.008692444198661381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,51200,0.05155822303560045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,51200,0.07933155695597331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,65536,0.07326577769385444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,12288,0.03056710958480835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,16384,0.021167111065652635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,12288,0.024961777859263953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,768,0.00903911143541336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,6144,0.01789422167672051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,16384,0.03005689051416185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,16384,0.03250044584274292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,10240,0.02375288969940609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,8192,0.021927111678653296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,12288,0.02664622167746226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,16384,0.038132445679770574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,8192,0.01908266709910499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,7168,0.020059555768966675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,8192,0.02100444502300686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,7168,0.012813333008024426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,6144,0.017835555805100333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,10240,0.01548266741964552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,6144,0.016719111137919955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,10240,0.0258879992696974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,5120,0.01584533353646596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,6144,0.018226666582955253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,8192,0.013813333378897773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,10240,0.02132266594303979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,5120,0.01073688848151101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,5120,0.01686755485004849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,7168,0.01940799918439653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,4096,0.009368889033794403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,4096,0.012675555215941535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,4096,0.015192889504962497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,12288,0.01757066614098019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,6144,0.011896000140243106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,3584,0.008834666675991481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,3072,0.009383999639087254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,3584,0.014501333236694336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,7168,0.017253332667880587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,2560,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,3584,0.01036800030204985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,2560,0.009360888765917884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,2560,0.007646222081449296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,3584,0.011075555450386472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,2560,0.012784000072214337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,5120,0.013843555417325763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,3072,0.01348888874053955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,2048,0.011418666276666852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,2048,0.007015110717879401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,1536,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,1536,0.009875555833180746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,1024,0.0042568888101312844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,4096,0.01221244451072481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,1536,0.006258666515350342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,1024,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,1024,0.009033777647548253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,2048,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,768,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,768,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,3072,0.010414222048388587
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,512,0.0031911111954185697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,768,0.009339555270142024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,512,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,256,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,512,0.00832799987660514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,1536,0.0070159998204973005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,256,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,256,0.005964444329341252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,128,0.002847111059559716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,2048,0.008369777765538957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,128,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,64,0.002557333351837264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,128,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,32,0.002887111157178879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,64,0.005296000176005893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,256,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,32,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,65536,0.10676888624827068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,3072,0.008193777667151557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,65536,0.06651644574271308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,51200,0.08435555299123128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,128,0.004951999833186467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,512,0.005684444473849402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,65536,0.09883644183476765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,51200,0.05042666527960035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,16384,0.03329688972897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,16384,0.020319110817379422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,16384,0.028373334142896865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,12288,0.0273333340883255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,16384,0.032271110349231295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,1024,0.006312888943486744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,12288,0.023519999451107446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,12288,0.026064000195927087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,10240,0.020611556039916146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,10240,0.023355555203225877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,10240,0.01484444406297472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,51200,0.07924089166853163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,8192,0.018238221605618794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,12288,0.01685422162214915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,65536,0.06127022372351753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,7168,0.018211555149820115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,8192,0.013439999686347114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,8192,0.020447111792034574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,51200,0.04928266670968798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,7168,0.016176889340082806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,7168,0.012423999607563019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,8192,0.020239111449983384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,6144,0.011414222419261932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,6144,0.014701333310869006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,6144,0.017641777793566387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,10240,0.02314844396379259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,5120,0.013552889227867126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,4096,0.01074577785200543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,768,0.0038773334688610504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,6144,0.016360888878504436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,7168,0.01908888916174571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,4096,0.01198755535814497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,4096,0.008610666626029545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,4096,0.014904000692897372
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,5120,0.014869333969222175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,3584,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,3584,0.014217777384652032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,5120,0.009719111025333405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,3072,0.010028444230556488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,5120,0.01601777805222405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,3072,0.007696000238259633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,3584,0.009512888888518015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,2560,0.008029333419269985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,2048,0.00700800038046307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,2560,0.012438221938080259
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,3072,0.008660444782839881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,2048,0.0069875551594628235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,2048,0.007889777421951294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,2048,0.01073333372672399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,1536,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,2560,0.007627555893527136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,1536,0.010055999788973067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,3072,0.01315111087428199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,1024,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,768,0.003546666767862108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,1024,0.009708444277445475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,1024,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,3584,0.010479110810491772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,512,0.003528000166018804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,768,0.009191110730171204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,1536,0.006705777926577463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,512,0.005626666463083691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,256,0.0032231110251612137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,512,0.00601244428091579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,2560,0.007328888608349695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,256,0.005249777601824866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,256,0.0057911111248864075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,128,0.0028968888024489083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,256,0.008349333372380998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,1536,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,64,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,128,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,128,0.004933333231343163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,128,0.008401778009202745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,1024,0.003930666794379552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,64,0.0052471111218134565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,65536,0.09055733018451267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,32,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,51200,0.06970844666163127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,65536,0.06001955270767212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,65536,0.0980844431453281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,768,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,768,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,32,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,51200,0.04733333322736952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,16384,0.030513776673210993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,12288,0.0214044451713562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,16384,0.031594667169782854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,65536,0.058865778976016574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,12288,0.021640888518757288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,12288,0.016214221715927124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,51200,0.0462026662296719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,16384,0.027267555395762127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,51200,0.07889866828918457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,10240,0.01462044484085507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,10240,0.023001778456899855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,16384,0.01980266637272305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,512,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,7168,0.014800889624489678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,8192,0.012822222378518848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,8192,0.02034133341577318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,10240,0.018720888429217868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,7168,0.011692444483439127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,6144,0.013796444568369122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,7168,0.018876444962289598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,6144,0.01089511149459415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,8192,0.016172443826993305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,6144,0.017550221747822232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,8192,0.016189333465364244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,5120,0.013088888592190213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,4096,0.011026666396194033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,5120,0.016200888488027785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,7168,0.015525332755512662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,12288,0.02613777750068241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,4096,0.01162755572133594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,3584,0.011396444506115384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,10240,0.01963200006220076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,6144,0.013790222505728403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,3584,0.008005333443482717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,3072,0.01037599974208408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,3584,0.013825777504179211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,5120,0.009711999860074785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,3072,0.008636444807052612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,3072,0.007669332954618666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,3072,0.012454222473833295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,4096,0.008375999828179678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,5120,0.01205244412024816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,2560,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,2560,0.011742221812407175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,3584,0.009027555584907532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,2048,0.0069857777820693115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,1536,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,2048,0.010763555765151978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,2560,0.00923466682434082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,1536,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,1024,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,4096,0.014766222900814481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,2048,0.008008889026112026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,768,0.004266666869322459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,1024,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,2048,0.007322666545708974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,768,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,512,0.0038577777643998465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,1536,0.00666844430896971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,512,0.005591111050711737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,2560,0.007673777639865875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,256,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,512,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,1024,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,512,0.008189332981904348
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,128,0.0028497777465316984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,1024,0.009396444592210982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,256,0.005157333281305102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,256,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,768,0.006275555739800136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,128,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,32,0.002879111096262932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,128,0.008352888955010308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,1536,0.010082667072614035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,64,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,32,0.004931555440028508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,65536,0.05704800287882487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,65536,0.05789688560697767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,51200,0.06403377983305189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,65536,0.09831466939714219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,768,0.008767999708652496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,51200,0.04439555605252584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,16384,0.02589955594804552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,51200,0.04656889041264852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,51200,0.07866399817996554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,12288,0.019997333486874897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,16384,0.024471999870406255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,16384,0.03151911165979173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,12288,0.0210844443904029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,12288,0.015997333659066092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,12288,0.025901334153281316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,64,0.0025395556456512874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,10240,0.018570666511853535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,8192,0.01568266749382019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,10240,0.023170666562186346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,65536,0.07914577590094672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,8192,0.016509332590632968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,8192,0.012477333347002665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,7168,0.013750221994188098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,10240,0.017196445001496207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,256,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,6144,0.012808000048001608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,7168,0.014977777997652689
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,7168,0.011434666812419891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,7168,0.018619555566045973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,5120,0.011160000330872007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,6144,0.009747555686367882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,10240,0.014476444986131458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,16384,0.01926488843229082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,5120,0.009023110899660323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,5120,0.011755555868148804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,5120,0.015998222761683993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,8192,0.020607999629444547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,3584,0.011060444845093621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,4096,0.009422221945391761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,4096,0.013789333403110504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,3584,0.008664000365469191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,3072,0.010127999716334874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,3584,0.013220444321632385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,6144,0.014352889524565803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,6144,0.01777866648303138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,3072,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,2560,0.008591999610265097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,3072,0.012091555529170565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,2560,0.007544000115659501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,2048,0.008316444853941599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,3584,0.007839110990365347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,128,0.004918222212129169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,2048,0.007332444190979004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,2048,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,4096,0.01074399964676963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,2048,0.010402667025725046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,2560,0.007343110938866933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,2560,0.011447110937701331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,1536,0.010060444474220276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,1024,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,1536,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,1024,0.0063324446479479475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,768,0.004206222378545337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,1024,0.009478222164842818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,3072,0.008667555948098501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,768,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,768,0.0061048890153567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,512,0.0038844446341196695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,768,0.009073778159088558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,4096,0.008556444611814287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,256,0.0032195556494924757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,512,0.005982222242487802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,512,0.005306666509972678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,512,0.008665777742862701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,128,0.002867555452717675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,256,0.005123555660247803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,1024,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,256,0.008534222013420528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,64,0.002862222285734283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,128,0.005623999983072281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,128,0.004898666507667965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,32,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,128,0.008045333127180735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,64,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,1536,0.008252444366614023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,32,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,51200,0.05589422252443102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,65536,0.05693066782421536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,256,0.005620444400442972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,65536,0.09825688600540161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,51200,0.045684443579779736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,51200,0.07862044705284967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,16384,0.024705777565638225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,16384,0.019173334042231243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,16384,0.031503111124038696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,65536,0.054233776198493115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,12288,0.025952888859642878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,12288,0.015878222054905362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,10240,0.015272888872358533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,16384,0.021911111142900255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,1536,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,10240,0.016393777396943834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,10240,0.02331022255950504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,12288,0.01756977703836229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,8192,0.014356444279352823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,8192,0.011769777370823754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,7168,0.012479110724396177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,12288,0.019142222073343065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,7168,0.010411555568377176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,7168,0.01349955548842748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,7168,0.018952000472280715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,10240,0.013844444519943662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,8192,0.01369511087735494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,5120,0.010635555618339114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,6144,0.01276088919904497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,5120,0.010650667051474253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,8192,0.0203448881705602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,4096,0.009506666825877296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,5120,0.008997333546479544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,6144,0.011757333245542316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,4096,0.010036444498433007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,3584,0.009355555805895064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,4096,0.013793778088357715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,65536,0.06843822532229953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,6144,0.017383111847771537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,3072,0.009109333157539368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,3584,0.008029333419269985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,3584,0.01311911145846049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,5120,0.015917332635985482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,3072,0.007672000262472365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,2560,0.007633777956167857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,51200,0.04276177618238661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,2560,0.007374222079912822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,2560,0.006800888727108638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,2048,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,2560,0.01145155562294854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,6144,0.009716444545321995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,1536,0.005971555494599872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,2048,0.006994666324721442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,2048,0.006618666566080517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,2048,0.010378667049937779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,1024,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,1536,0.006227555374304454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,3072,0.007696888513035244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,1536,0.010037333601050908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,3072,0.012092444631788464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,1024,0.006310222049554189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,4096,0.0083155557513237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,768,0.005711110929648082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,768,0.00609688874748018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,512,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,768,0.008930666579140557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,3584,0.007855110698276097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,512,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,1536,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,512,0.00870666652917862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,768,0.004552000098758274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,1024,0.005561777700980504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,256,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,128,0.0032071111102898917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,256,0.005648889061477449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,128,0.00564444437623024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,64,0.002895111011134254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,128,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,32,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,512,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,256,0.003552888830502828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,65536,0.06103377872043186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,32,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,65536,0.05187466740608215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,256,0.008038221961922115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,51200,0.048116445541381836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,65536,0.09781955348120795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,128,0.004907555464241239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,51200,0.04121066795455085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,16384,0.01992177797688378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,51200,0.0785822206073337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,64,0.00508266646001074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,12288,0.015503111812803479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,16384,0.03122488988770379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,1024,0.009183111290136972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,65536,0.05608089102639092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,12288,0.017870222528775532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,12288,0.025735111700163946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,51200,0.045259555180867515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,10240,0.01579466627703773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,8192,0.012294222083356647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,10240,0.02329066726896498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,8192,0.01185955521133211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,8192,0.013842666314707862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,7168,0.01124622258875105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,8192,0.020225778222084045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,7168,0.010672888822025724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,7168,0.012691555751694573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,6144,0.010538666612572141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,7168,0.01889155473974016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,6144,0.009697777529557547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,5120,0.009833777944246927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,16384,0.021104888783560857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,5120,0.00945688866906696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,10240,0.014304889572991265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,12288,0.015855110353893705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,5120,0.015166223049163818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,10240,0.013595555391576556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,6144,0.017239999439981248
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,4096,0.013431111143694984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,3584,0.0086666668454806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,16384,0.018744000130229525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,3584,0.007687110867765214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,3584,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,3072,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,5120,0.00906133320596483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,3072,0.007714666426181793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,2560,0.0069955554273393415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,6144,0.011075555450386472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,4096,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,2560,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,2048,0.006085333310895496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,3584,0.01276088919904497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,4096,0.009288888838556077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,3072,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,3072,0.012079111403889127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,1536,0.005984000033802456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,1536,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,2560,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,1536,0.010031999813185798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,1024,0.004587555511130227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,1536,0.0069439998931354946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,2048,0.00665511108107037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,1024,0.0064426664676931165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,768,0.004601777841647466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,1024,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,2048,0.010911111202504901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,512,0.0035964445107513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,768,0.009290667043791877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,2560,0.011063111325105032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,512,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,256,0.003198222153716617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,1024,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,2048,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,4096,0.008973333570692275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,256,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,256,0.008347555167145198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,512,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,128,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,64,0.0025128889828920364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,128,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,512,0.00871377769443724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,32,0.002534222271707323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,65536,0.05102755626042684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,64,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,32,0.005017777697907554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,256,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,768,0.006260444306664997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,65536,0.04993155598640442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,51200,0.04079733292261759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,65536,0.055445333321889244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,65536,0.09817866484324138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,51200,0.04502044452561272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,16384,0.016125332978036668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,128,0.005942222144868638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,51200,0.07844533522923787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,12288,0.01312888910373052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,16384,0.01827822294500139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,16384,0.02033155494266086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,16384,0.03141599893569946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,10240,0.011729777687125735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,768,0.005333333379692501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,12288,0.025586666332350835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,10240,0.013835555149449242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,8192,0.011101333631409539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,8192,0.012010667059156628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,128,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,8192,0.011083555718262991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,7168,0.01072799993885888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,10240,0.01313511116637124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,10240,0.022689777943823073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,6144,0.009979555176364051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,7168,0.010527999864684211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,7168,0.018556444181336295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,6144,0.009770666559537252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,12288,0.014490667316648694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,5120,0.008887999587588841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,8192,0.01998044384850396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,5120,0.008408000071843466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,4096,0.010216000179449717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,7168,0.01071733319097095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,4096,0.007672000262472365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,4096,0.008218666745556725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,12288,0.015200000670221118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,3584,0.009359999663299983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,3584,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,3584,0.012418666647540199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,3072,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,3584,0.008041777544551426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,3072,0.007231999602582719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,3072,0.012092444631788464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,2560,0.0059262220230367445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,2560,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,5120,0.015383111106024848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,2560,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,2048,0.006611555400821898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,6144,0.00903200027015474
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,4096,0.013482666677898832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,2048,0.006312888943486744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,1536,0.005592888842026393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,2048,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,2048,0.010451555252075195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,5120,0.009024889104896123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,1536,0.00591555568906996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,1024,0.00462311092350218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,1536,0.009709333380063375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,2560,0.011403555671374003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,768,0.004219555606444676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,1024,0.009349333743254343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,6144,0.017238222890430026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,768,0.005264889034960005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,768,0.0063182223174307085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,768,0.00924622184700436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,51200,0.03980355461438497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,1536,0.0067528887755341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,512,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,256,0.0031999999450312722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,512,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,512,0.008610666626029545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,1024,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,256,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,256,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,256,0.007993777592976889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,64,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,128,0.005722666780153911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,128,0.004908444566859139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,512,0.0035626664757728577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,128,0.00776622196038564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,65536,0.037676443656285606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,64,0.004922666483455234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,32,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,51200,0.03050044510099623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,65536,0.09781599707073635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,1024,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,51200,0.03872444563441806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,16384,0.013667555318938361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,32,0.0026648888985315957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,51200,0.07841244671079847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,16384,0.0177182224061754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,16384,0.017877333694034152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,16384,0.031217777066760596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,65536,0.05491022268931071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,12288,0.014597333139843412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,128,0.002895111011134254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,3072,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,51200,0.044241779380374484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,10240,0.013112000293201871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,10240,0.012128888732857175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,10240,0.022421333524915908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,8192,0.010415111151006486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,65536,0.04787733488612705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,8192,0.010032888915803697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,10240,0.010344888601038191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,8192,0.019581332802772522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,12288,0.025440888272391424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,7168,0.010443555812040964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,6144,0.009289777941173976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,7168,0.01752088963985443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,12288,0.011490667031870948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,6144,0.00886933339966668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,6144,0.009482666850090027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,5120,0.008045333127180735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,6144,0.01639111174477471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,12288,0.013567110730542077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,7168,0.009681777821646797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,5120,0.009018667042255402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,5120,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,5120,0.014847111370828418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,7168,0.009348444640636444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,4096,0.01350933313369751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,3584,0.007324444750944774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,3072,0.006326222171386083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,3584,0.012768000364303589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,8192,0.010755555497275459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,4096,0.007299555672539606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,2560,0.005658666706747479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,3072,0.007320889168315464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,3072,0.012104888757069906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,3584,0.006992888947327931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,2048,0.006315555423498154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,2560,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,2560,0.011394666300879585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,2048,0.006716444260544247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,1536,0.0053102220926019884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,3072,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,4096,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,2560,0.007373332977294922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,1024,0.0042444442709287005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,2048,0.0063075555695427795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,2048,0.0107004443804423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,1024,0.005665777872006099
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,768,0.00387999994887246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,1024,0.006292444550328785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,1024,0.009361777868535783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,768,0.005588444570700328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,512,0.003583111282851961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,768,0.009012444151772393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,1536,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,512,0.0052826665341854095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,512,0.006114666660626729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,256,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,512,0.00832799987660514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,256,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,128,0.0028560000161329904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,1536,0.005975111077229182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,256,0.008399110701349046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,768,0.006269333263238271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,128,0.0057884446448749965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,64,0.002598222138153182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,32,0.002543111021320025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,64,0.005015111217896144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,3584,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,32,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,256,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,65536,0.04650311006440056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,51200,0.02234577801492479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,65536,0.05442400111092461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,65536,0.09761333465576172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,128,0.004958222309748332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,16384,0.014399111270904541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,51200,0.04404533240530226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,51200,0.07836800151401095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,65536,0.02678044471475813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,16384,0.016000888413853116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,12288,0.012816888590653738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,16384,0.015832000308566622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,12288,0.013391111459996967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,10240,0.011409777734014722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,4096,0.0074311114019817775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,1536,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,51200,0.03734133309788174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,10240,0.011076444553004371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,8192,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,10240,0.011773332953453064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,10240,0.02166755497455597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,7168,0.009680888719028896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,8192,0.01072622256146537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,8192,0.010051555931568146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,8192,0.018909333480728995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,6144,0.008381333616044786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,7168,0.010082667072614035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,12288,0.012424888710180918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,12288,0.025111999776628282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,5120,0.0079644446571668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,6144,0.009008888569143083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,128,0.007722666694058313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,6144,0.016200888488027785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,5120,0.008325333396593729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,5120,0.008763555851247575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,5120,0.014505777094099255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,16384,0.03083733386463589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,4096,0.007687110867765214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,4096,0.008303999900817871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,7168,0.01704088846842448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,6144,0.009570666485362584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,3584,0.0073520003093613525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,3584,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,3072,0.0063982220987478895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,3584,0.012416000167528788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,4096,0.007347555624114142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,2560,0.0059164443777667145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,3072,0.0073724447025193115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,3072,0.011784888803958893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,7168,0.009363555245929295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,2048,0.006270222365856171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,4096,0.01312622262371911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,2048,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,2048,0.010384000009960597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,1536,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,2048,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,3072,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,1536,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,1024,0.004556444370084339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,1536,0.010098666780524785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,2560,0.01107911103301578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,768,0.0038853333228164246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,1024,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,1024,0.005315555466545953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,1024,0.009353777600659264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,768,0.006270222365856171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,768,0.005280888742870755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,768,0.008742222355471717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,512,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,256,0.0031902222997612427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,512,0.008004444340864817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,2560,0.006613333192136552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,256,0.0058017778727743365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,128,0.0028924445311228433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,256,0.007660444411966536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,512,0.0038373333712418876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,128,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,128,0.005707555760939916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,64,0.002540444541308615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,32,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,512,0.005924444645643234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,768,65536,0.021760889225535925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,64,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,32,0.005003555367390315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,256,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,65536,0.04589777853753832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,65536,0.054125332170062594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,65536,0.09734311368730332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,51200,0.03700266612900628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,51200,0.04402577877044678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,51200,0.07839733362197876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,1536,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,3584,0.006689777804745569
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,768,12288,0.011400889191362592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,16384,0.030771556827757094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,768,51200,0.0184568895234002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,12288,0.012270222107569376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,12288,0.013441777891582914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,12288,0.02498133314980401
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,128,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,10240,0.011061333119869232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,10240,0.011749332977665795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,10240,0.022317333353890315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,16384,0.01588444411754608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,8192,0.009678222239017487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,768,7168,0.008412444757090675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,8192,0.018698667486508686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,768,10240,0.01016355554262797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,7168,0.009249777429633671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,7168,0.010048000348938836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,768,6144,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,7168,0.01755199995305803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,768,8192,0.009387555221716562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,768,5120,0.006953777538405524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,2560,0.007339555356237624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,6144,0.01551466683546702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,8192,0.010605333579911126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,768,4096,0.007320889168315464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,768,16384,0.01035733355416192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,4096,0.007363555332024892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,4096,0.007991111112965478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,16384,0.014490667316648694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,4096,0.013472889032628802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,5120,0.008119111259778341
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,3584,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,5120,0.014836443795098199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,768,3072,0.006745777610275481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,3584,0.01276266657643848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,3072,0.007633777956167857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,3072,0.006989333364698622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,768,2560,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,3072,0.012072000238630505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,6144,0.008779555559158325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,2560,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,768,2048,0.0052622221410274506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,2560,0.011415111521879831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,5120,0.00868088917599784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,2048,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,768,1536,0.0052791109515561005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,1536,0.00600000015563435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,6144,0.009680888719028896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,2560,0.007299555672539606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,1536,0.006682666639486949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,768,1024,0.004234666625658671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,3584,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,2048,0.01071111112833023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,768,768,0.003536888708670934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,1024,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,1024,0.009053332938088311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,768,0.006299555715587404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,768,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,768,512,0.0035546666218174826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,768,0.008722667064931657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,512,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,512,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,512,0.008468444148699442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,2048,0.006671111202902264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,768,128,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,256,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,256,0.008394666843944127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,128,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,768,64,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,128,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,768,32,0.0025155554629034465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,1536,0.010072888599501716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,768,256,0.003212444484233856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,512,65536,0.017491555876202054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,32,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,1024,0.005660444498062134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,65536,0.05166844526926676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,512,51200,0.015068444940778943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,65536,0.0970355537202623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,51200,0.036504000425338745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,768,3584,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,512,16384,0.010996444357766045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,51200,0.07826577954822116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,64,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,16384,0.014497778481907315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,16384,0.03016888764169481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,65536,0.045200887653562755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,12288,0.02472177810139126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,512,10240,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,51200,0.04045777850680881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,128,0.00814844419558843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,10240,0.011911110745535957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,10240,0.0110880004035102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,512,8192,0.008051555189821456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,512,12288,0.008678221868144141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,10240,0.021394666698243883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,512,7168,0.007684444387753804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,8192,0.01071111112833023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,8192,0.010054222411579555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,256,0.005588444570700328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,8192,0.018910222583346896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,512,6144,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,7168,0.010041777458455827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,12288,0.013467555244763693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,512,5120,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,6144,0.009701333112186855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,6144,0.008683555656009251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,6144,0.016195555528004963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,5120,0.009020444419648912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,5120,0.008279999925030602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,5120,0.014762666490342883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,4096,0.007354666789372762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,512,3584,0.006681777950790193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,7168,0.009213333328564962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,7168,0.016912889149453905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,3584,0.007326222128338284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,512,3072,0.006634666687912411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,3584,0.007874666816658443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,3584,0.012641777594884237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,16384,0.01590666671593984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,512,2560,0.006317333214812809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,3072,0.0069573331210348345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,3072,0.007670222057236566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,3072,0.011400000088744693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,4096,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,2560,0.007357333269384172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,12288,0.012129777835475074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,2560,0.011415111521879831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,2048,0.00664533343580034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,2048,0.006197333335876465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,2048,0.01075111081202825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,1536,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,1536,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,512,1024,0.003903999924659729
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,1536,0.009708444277445475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,1024,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,1024,0.005295111073387994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,4096,0.013445333474212222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,2560,0.00638755535085996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,512,1536,0.00434666665063964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,768,0.005294222384691238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,512,512,0.003245333416594399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,768,0.00609688874748018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,512,4096,0.007303111255168915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,512,0.004936000125275718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,512,256,0.0032071111102898917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,512,0.008350222474998897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,512,2048,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,1024,0.009022222624884712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,512,128,0.0028853333658642242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,256,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,256,0.007991111112965478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,128,0.005664000080691443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,768,0.009268444445398118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,512,64,0.002570666579736604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,512,32,0.0025111111915773815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,64,0.004967111266321606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,256,65536,0.01587377819750044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,512,768,0.003532444437344869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,256,0.004990222139490975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,65536,0.042528887589772545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,256,51200,0.01347555551264021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,65536,0.09637155797746445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,128,0.004599110947714912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,128,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,256,16384,0.008658666577604082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,51200,0.03776355584462484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,51200,0.03464622298876444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,512,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,51200,0.07736977603700426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,16384,0.015848888291252982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,16384,0.029729776912265356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,12288,0.01313777764638265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,256,10240,0.007660444411966536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,12288,0.024384000235133704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,65536,0.04691822330156962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,10240,0.012096000214417776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,10240,0.021306667062971327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,8192,0.009872888525327047
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,32,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,256,7168,0.006959999601046245
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,12288,0.012113778127564324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,7168,0.00928177767329746
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,10240,0.010789333118332757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,256,6144,0.0062977779242727495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,7168,0.017266665895779926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,256,12288,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,6144,0.008618666893906063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,16384,0.01419911119672987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,6144,0.016180444094869826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,8192,0.01073333372672399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,5120,0.009010666774378883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,5120,0.0079644446571668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,5120,0.014688889185587565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,256,8192,0.006990222467316522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,7168,0.010029333333174387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,4096,0.006989333364698622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,256,3584,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,4096,0.01344622257683012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,8192,0.018851555056042142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,3584,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,256,3072,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,3584,0.012508444488048553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,3072,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,256,4096,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,256,2560,0.006329777754015393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,3072,0.011786667009194693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,2560,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,4096,0.008328888979223039
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,2560,0.011423110961914062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,256,5120,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,2048,0.006311999840868845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,6144,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,256,1536,0.004270222038030624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,2048,0.006637333167923822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,1536,0.0064702220261096954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,1536,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,1536,0.010417777631017896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,2560,0.006304889089531369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,256,2048,0.00592533333433999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,1024,0.005653333332803514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,256,768,0.003526222374704149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,1024,0.00648266656531228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,1024,0.009032888544930352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,256,512,0.0032542221662071017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,768,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,768,0.005614222337802251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,2048,0.010381333529949188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,768,0.0086666668454806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,512,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,3072,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,256,1024,0.003847111016511917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,256,128,0.0028782222006056043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,256,0.004935111022657818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,256,0.008394666843944127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,128,0.0046844445168972015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,128,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,256,64,0.0025235555238193935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,256,256,0.0032053333189752367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,256,32,0.0028817777832349143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,512,0.0052782222628593445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,128,65536,0.014382221632533602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,512,0.00832799987660514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,256,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,32,0.0046471113132105935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,128,51200,0.012792888614866467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,65536,0.04594844579696655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,65536,0.09582400321960449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,128,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,3584,0.007041778001520369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,128,16384,0.007648888561460707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,51200,0.03381422162055969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,51200,0.03735911183887058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,128,12288,0.007667555577225155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,51200,0.076127999358707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,16384,0.014131555954615274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,16384,0.029671112696329754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,128,10240,0.007127111156781514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,12288,0.012098666694429187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,12288,0.023823110593689814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,65536,0.042111999458736844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,10240,0.012080888781282636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,10240,0.021578666236665513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,8192,0.009688888986905416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,16384,0.015856888559129503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,128,7168,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,8192,0.010834666589895884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,12288,0.013117333253224691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,7168,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,128,6144,0.006305777778228124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,10240,0.011048888994587792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,128,8192,0.006674666785531574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,6144,0.009407111340098912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,6144,0.008676444490750631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,6144,0.015838222371207345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,8192,0.018651554981867473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,5120,0.008010666403505538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,5120,0.014478221535682678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,7168,0.010045333041085137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,7168,0.01720177796151903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,4096,0.013429332938459186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,4096,0.008001777860853408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,128,5120,0.005990222096443176
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,3584,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,3584,0.007970666719807519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,128,3072,0.005647111270162795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,4096,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,3072,0.012084444363911947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,128,3584,0.005971555494599872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,3072,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,2560,0.006664888726340399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,2560,0.011056889262464313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,128,2048,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,3584,0.012169777519173093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,2048,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,2048,0.010394666757848527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,128,2560,0.006292444550328785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,128,1536,0.0042239998777707415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,64,0.004984000076850255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,1536,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,128,1024,0.00388711111413108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,1024,0.005326222214433882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,2048,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,1024,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,128,768,0.003567999849716822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,128,4096,0.005624889085690181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,768,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,768,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,128,512,0.0031946665710873077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,1024,0.00869955536392
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,1536,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,512,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,128,256,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,512,0.008350222474998897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,3072,0.0070088886552386815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,768,0.008976889153321585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,128,128,0.0028799999919202593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,256,0.00564533347884814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,512,0.005179555465777715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,128,0.004923555586073133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,128,64,0.002546666603949335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,128,0.005623111294375525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,5120,0.009018667042255402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,64,0.004939555707905027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,32,0.004895110925038655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,256,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,64,51200,0.009741333623727163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,256,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,64,16384,0.0069679998689227635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,65536,0.041787554820378624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,64,12288,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,128,0.007000000112586551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,2560,0.007515555454625024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,12288,0.01181333346499337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,64,8192,0.006269333263238271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,64,65536,0.010391111175219217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,64,7168,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,8192,0.009679111341635386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,7168,0.009190222455395592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,51200,0.03360088997417026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,16384,0.014161777165200977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,1536,0.006672888994216919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,6144,0.008661333057615492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,64,4096,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,10240,0.011061333119869232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,64,3584,0.006452444526884291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,4096,0.007345777418878343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,64,3072,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,3584,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,64,6144,0.006321777900060018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,64,10240,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,64,2048,0.00565422202150027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,2560,0.00628266649113761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,64,1536,0.00423911131090588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,5120,0.008018666671382057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,64,1024,0.0038933331767717996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,1536,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,64,768,0.0035386664999855887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,1024,0.005648889061477449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,64,2560,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,64,512,0.003175111073586676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,64,256,0.002887111157178879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,128,32,0.002577777744995223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,64,128,0.0025457777082920074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,512,0.0052426668504873914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,64,64,0.0025191110455327565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,64,32,0.002526222210791376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,128,0.004937777916590373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,64,0.00490399988161193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,32,0.004912000149488449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,32,51200,0.008389333056079017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,65536,0.04178933302561442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,32,16384,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,768,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,32,12288,0.006299555715587404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,256,0.004962666581074397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,32,10240,0.006157333238257303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,2048,0.006329777754015393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,32,65536,0.008728889127572378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,10240,0.010774222513039907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,32,7168,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,8192,0.009696889254781935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,32,6144,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,64,5120,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,32,5120,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,6144,0.008301333420806462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,32,4096,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,12288,0.012057777908113269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,32,8192,0.005631999837027655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,32,3584,0.005653333332803514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,32,3072,0.0058062221441004015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,4096,0.006992888947327931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,32,2560,0.00562755556570159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,3584,0.00702044450574451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,32,2048,0.005660444498062134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,3072,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,32,1536,0.004571555389298333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,2560,0.006255110932721033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,32,1024,0.0038408889538711975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,16384,0.01447111037042406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,32,768,0.003559111307064692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,1536,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,5120,0.007650666766696506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,1024,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,32,256,0.002536888958679305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,768,0.005281777845488654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,32,128,0.002581333327624533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,512,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,32,64,0.0025048889219760895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,256,0.004943111290534337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,32,32,0.002528888897763358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,128,0.004936000125275718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,7168,0.00904888908068339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,64,0.004600888739029567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,2048,0.006261333409282897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,32,0.004940444396601783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,4,32,512,0.0031991110493739447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,16384,0.33682576815287274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,51200,0.48413244883219403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,12288,0.2595573266347249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,16384,0.13662577999962702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,10240,0.21724355220794678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,12288,0.12154311603969997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,51200,0.43886666827731663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,10240,0.12966488467322454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,8192,0.17716977331373426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,10240,0.10499822431140476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,51200,0.0335804455810123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,8192,0.11768799357944065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,7168,0.14621244536505804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,8192,0.07297066847483318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,12288,0.194961773024665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,6144,0.127966218524509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,7168,0.06307911210589938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,7168,0.1117031176884969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,5120,0.10951022307078044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,6144,0.08547200096978082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,6144,0.0540888872411516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,4096,0.09156711233986749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,5120,0.04677333434422811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,16384,0.1808711157904731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,3584,0.07291999790403578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,4096,0.06574666500091553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,4096,0.039417776796552874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,3072,0.06335200203789605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,3584,0.03735911183887058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,2560,0.05319199959437052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,3072,0.036729776197009616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,3072,0.047135111358430654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,2048,0.04351822204060025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,2560,0.033591111501057945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,2048,0.02897955642806159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,5120,0.07358044385910034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,1536,0.0335804455810123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,51200,1.0139457914564345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,1536,0.02368533280160692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,1536,0.024663110574086506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,3584,0.05901333358552721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,1024,0.023403555154800415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,1024,0.01828444500764211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,1024,0.020271999968422782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,512,0.01366311146153344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,2560,0.043120000097486705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,768,0.01690933273898231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,256,0.008653333617581261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,2048,0.03702222307523092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,512,0.013064888616402945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,3072,0.00666133314371109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,256,0.010759111079904767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,256,0.00997155573632982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,64,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,128,0.010732444624106089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,128,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,64,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,768,0.015537778536478678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,32,0.008711111214425828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,512,0.013047999805874295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,128,0.006441777778996362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,65536,0.5126630995008681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,768,0.018696889281272888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,65536,0.9740453296237521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,16384,0.2507377730475532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,16384,0.14640622668796116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,65536,0.42513510915968156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,51200,0.35446667671203613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,16384,0.11167821619245742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,12288,0.14660355779859754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,51200,0.7540151278177897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,12288,0.08217155271106295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,8192,0.13138666417863634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,10240,0.12440178129408096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,10240,0.07059733072916667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,7168,0.11577778392367893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,12288,0.19194489055209688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,8192,0.09262222051620483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,6144,0.10078844759199355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,7168,0.0774515536096361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,8192,0.0563875569237603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,5120,0.08935733636220296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,32,0.005624889085690181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,51200,0.38776977856953937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,4096,0.07159911261664496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,6144,0.04462488823466831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,5120,0.03866133424970839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,4096,0.053786665201187134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,4096,0.032834665642844305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,3072,0.05540000067816841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,3584,0.03130133284462823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,3584,0.0492266681459215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,7168,0.052899556027518384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,3072,0.029488888051774766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,10240,0.16122578250037298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,3072,0.03849155704180399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,2048,0.0377786656220754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,2560,0.035951998498704694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,1536,0.029713776376512315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,2048,0.024381332927280005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,3584,0.06400711006588407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,1024,0.021524444222450256
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,1536,0.019985778464211356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,1536,0.020272889071040683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,768,0.016168889072206285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,1024,0.015538665983412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,1024,0.017608000172509086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,512,0.011436444189813403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,768,0.013555555707878537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,6144,0.06874577866660224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,768,0.014567999376191033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,256,0.007297777467303806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,512,0.011249778171380361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,128,0.003923555629120933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,512,0.011416888899273343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,256,0.009693333672152625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,64,0.0035457776652442086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,256,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,128,0.009471110999584198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,32,0.003528888854715559
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,128,0.007334222396214803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,64,0.007771555748250749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,32,0.008040000167157915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,2560,0.04723822077115377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,2560,0.02677244444688161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,51200,0.2496133380466037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,65536,0.1743546724319458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,16384,0.08567111359702216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,2048,0.030772444274690416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,51200,0.13031644291347927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,16384,0.04228355487187704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,16384,0.06399555338753594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,65536,0.32089421484205455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,12288,0.03314044409328037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,12288,0.05416977736685011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,10240,0.034640888373057045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,10240,0.04314400090111626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,5120,0.060488886303371854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,8192,0.024111110303137038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,12288,0.0665644473499722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,7168,0.03803999887572394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,7168,0.02200444373819563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,10240,0.05635999970965915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,7168,0.0351839992735121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,6144,0.01991911066903008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,65536,0.15233866373697916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,8192,0.043050666650136314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,4096,0.023708444502618577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,5120,0.0243857784403695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,51200,0.1259253289964464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,3584,0.020624000165197585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,4096,0.023002665903833177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,6144,0.033062222931120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,3072,0.018209778600268893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,3584,0.019523556033770244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,6144,0.029319110843870375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,2560,0.015480000111791821
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,3072,0.015499555402331881
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,2560,0.014762666490342883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,2048,0.01311911145846049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,5120,0.017737777696715463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,2048,0.013227555486891003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,1536,0.010799110763602786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,5120,0.02769688930776384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,1536,0.010820444259378644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,1536,0.01092888911565145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,8192,0.03666844301753574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,2560,0.016145777371194627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,4096,0.015832889411184523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,768,0.0068959999415609576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,1024,0.009690666364298927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,3584,0.015639111399650574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,512,0.005162666655249066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,768,0.007636444436179266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,256,0.0034737777378824023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,512,0.0076453329788313965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,512,0.0070062221752272705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,128,0.0032248888164758682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,1024,0.008671999805503422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,256,0.005897777775923411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,64,0.0028479999552170434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,128,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,128,0.00553688903649648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,32,0.0028853333658642242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,1024,0.009150221943855286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,64,0.005592000153329637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,768,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,65536,0.24396443367004395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,51200,0.18858933448791504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,65536,0.1414239936404758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,256,0.0069982219073507525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,16384,0.06594844659169515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,51200,0.1031893359290229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,16384,0.037455999188952975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,16384,0.05776800049675835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,2048,0.01367644468943278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,32,0.005991999887757831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,12288,0.04343644446796841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,65536,0.14589777257707384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,8192,0.038528889417648315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,10240,0.036814222733179726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,8192,0.022295110755496558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,8192,0.03196977906756931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,7168,0.03478399912516276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,12288,0.02970577942000495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,6144,0.03038755390379164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,7168,0.020006222857369315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,6144,0.018251554833518136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,6144,0.025275554921891954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,51200,0.11137510670555963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,5120,0.016241777274343703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,5120,0.021527111530303955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,12288,0.051903112067116626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,4096,0.014819555812411837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,4096,0.019698666201697457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,10240,0.04348088966475593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,3584,0.014135110709402295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,3584,0.01737333337465922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,4096,0.02330222229162852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,3072,0.014807111687130399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,2560,0.015625778171751235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,3072,0.01331022216214074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,3584,0.02021511064635383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,7168,0.028169777658250596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,2048,0.013798221945762634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,2560,0.012792000340090858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,3072,0.018411555223994784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,3072,0.017907554904619854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,2048,0.011392888923486074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,2048,0.012747555143303342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,1024,0.00702400008837382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,1536,0.010033778018421596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,1536,0.010023111270533668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,768,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,1024,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,768,0.007700444095664554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,512,0.0038897775941424896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,2560,0.014840889308187695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,10240,0.02922933300336202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,256,0.0032675556010670135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,512,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,512,0.006636444479227066
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,128,0.0029200000895394217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,256,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,256,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,64,0.0028782222006056043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,1024,0.008653333617581261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,32,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,64,0.005595555735958948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,768,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,32,0.005710222240951326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,1536,0.008988444175985124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,51200,0.16135288609398737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,65536,0.14123733838399252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,65536,0.1274720033009847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,16384,0.05766399701436361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,128,0.006694222076071634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,51200,0.09863555431365967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,128,0.005488889084921942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,12288,0.04496888981925117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,16384,0.047854221529430814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,16384,0.035079111655553184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,12288,0.028456889920764502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,12288,0.037969777981440224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,10240,0.024142222272025213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,5120,0.026899556318918865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,7168,0.02932000160217285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,8192,0.02077866593996684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,8192,0.027943111128277246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,6144,0.0262044436401791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,51200,0.10691466596391465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,7168,0.0233608881632487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,5120,0.022846221923828125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,6144,0.017509332961506315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,6144,0.020995555652512446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,10240,0.038712888956069946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,4096,0.019340443942281935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,5120,0.01573066744539473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,8192,0.032039112514919706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,4096,0.014156444205178155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,65536,0.2040168841679891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,4096,0.01684977776474423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,3584,0.014875556031862894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,7168,0.018749333090252347
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,2560,0.014152000347773233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,3072,0.013456000222100152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,2560,0.011439111497667102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,2048,0.01112711098459032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,5120,0.01883555617597368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,10240,0.033518221643235944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,2048,0.010477333433098264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,3072,0.016153777639071148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,1536,0.009033777647548253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,1024,0.006312888943486744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,1024,0.007312888900438945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,3584,0.017631111873520743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,1024,0.007732444339328342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,768,0.004864000197913912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,1536,0.008884444832801819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,512,0.003556444413132138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,768,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,1536,0.00904888908068339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,512,0.006647999915811751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,512,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,256,0.006674666785531574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,2560,0.012378666963842181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,128,0.002830222249031067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,128,0.006903111106819577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,128,0.0052879999081293745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,64,0.002570666579736604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,768,0.007308444215191736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,32,0.0028488888508743713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,65536,0.16917599572075737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,256,0.0032035555276605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,32,0.0053635554181204895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,51200,0.12815733750661215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,65536,0.08908977773454453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,65536,0.07098844316270617
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,16384,0.046498666206995644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,256,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,2048,0.011434666812419891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,51200,0.05747644768820869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,12288,0.03343910972277323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,16384,0.022658665974934895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,16384,0.034521778424580894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,64,0.00563733321097162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,12288,0.018624888526068795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,12288,0.02886755598915948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,8192,0.023802666200531855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,10240,0.023704888092146978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,8192,0.014769777655601501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,8192,0.020833777056799993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,7168,0.020663999848895602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,51200,0.06837422317928739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,6144,0.019124444988038804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,7168,0.013793778088357715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,7168,0.018904000520706177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,10240,0.028103109863069322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,5120,0.015875554747051664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,6144,0.016532444291644625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,4096,0.013988445202509562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,10240,0.01755199995305803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,5120,0.015224888920783997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,3584,0.012068444656001197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,4096,0.013444444371594323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,3584,0.013457777599493662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,3072,0.01108000013563368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,3584,0.012503999802801343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,2560,0.009901333186361525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,3072,0.009789333575301701
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,3072,0.012746666868527731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,6144,0.012703110774358114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,2048,0.008585777547624376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,2560,0.009016888837019602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,2560,0.010758221977286868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,5120,0.011910222470760345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,2048,0.00833333366447025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,1024,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,2048,0.009365333451165093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,4096,0.010751999914646149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,768,0.003950222084919612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,1536,0.007351111206743453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,1024,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,512,0.003546666767862108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,1024,0.006646222124497096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,768,0.00666844430896971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,3584,0.010382222632567087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,256,0.0031742221779293488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,768,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,128,0.0029191111938820947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,512,0.005622222191757626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,256,0.005280888742870755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,128,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,64,0.0026808888134029177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,128,0.004935111022657818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,1536,0.007690666450394525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,64,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,65536,0.14766311645507812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,512,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,65536,0.06607821914884779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,3072,0.011051555474599203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,51200,0.1162311103608873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,16384,0.04211555586920845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,1536,0.007135111424658034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,32,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,51200,0.06202489137649536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,12288,0.034040000703599714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,16384,0.03355111016167535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,10240,0.02896889050801595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,12288,0.028174221515655518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,65536,0.08127378092871772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,10240,0.02273511058754391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,8192,0.025264889001846313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,8192,0.01423911088042789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,7168,0.018811555372344125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,16384,0.022096888886557683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,7168,0.013154666456911298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,7168,0.017272889614105225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,6144,0.01735111077626546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,6144,0.012183999849690331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,5120,0.015133332875039844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,8192,0.020614221692085266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,5120,0.014562666416168213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,4096,0.013448889056841532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,5120,0.011108444796668159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,32,0.005283555636803309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,4096,0.009766221874290043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,4096,0.012747555143303342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,10240,0.0158924443854226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,3584,0.009217778013812171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,3072,0.01035822182893753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,6144,0.016120000018013846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,2560,0.009010666774378883
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,3072,0.008367111285527548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,3072,0.01036000003417333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,2048,0.008040000167157915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,2560,0.010327999790509542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,12288,0.01789777808719211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,2048,0.008671999805503422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,256,0.005976000179847081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,3584,0.012095111111799875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,1536,0.007344000041484833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,51200,0.05352444450060526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,2560,0.00798044436507755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,1024,0.006693333387374878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,768,0.003583999971548716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,3584,0.011381333072980246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,768,0.006270222365856171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,512,0.0032151111712058387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,1536,0.007065777977307637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,512,0.006275555739800136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,256,0.0028888889484935333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,256,0.005240889059172736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,1536,0.006837333242098491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,128,0.0028808888875775864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,256,0.00591822216908137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,768,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,128,0.00592533333433999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,128,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,32,0.002579555536309878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,1024,0.004546666724814309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,64,0.005236444373925527
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,65536,0.12369955910576715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,32,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,1024,0.006696888970004187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,65536,0.07217778099907769
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,65536,0.06353777647018433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,16384,0.03795555565092299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,51200,0.05137422349717882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,51200,0.055918223328060575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,12288,0.029976000388463337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,16384,0.02132622235351139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,16384,0.029976887835396662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,10240,0.02639555599954393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,12288,0.017208000024159748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,8192,0.022091555926534865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,10240,0.015504888362354703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,10240,0.020947555700937908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,7168,0.01995822290579478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,2048,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,8192,0.02088622252146403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,6144,0.017893332574102614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,7168,0.012780444489585029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,7168,0.016685333516862657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,5120,0.015796444482273526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,6144,0.011821333732869891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,6144,0.015141333142916361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,4096,0.011916444533401065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,512,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,5120,0.013842666314707862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,3584,0.010404444403118556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,4096,0.009011555876996782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,4096,0.012456888953844706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,3584,0.008448888858159384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,3584,0.011315555208259158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,3072,0.008374222450786168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,64,0.0025635556214385563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,2560,0.00835111074977451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,5120,0.010605333579911126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,2560,0.007779555188284979
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,2048,0.007354666789372762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,12288,0.024534222152498033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,1536,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,2048,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,8192,0.013788444300492605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,2048,0.0084915558497111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,1536,0.006691555596060223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,1536,0.007010666860474481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,1024,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,1024,0.0063928887248039246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,3072,0.010013333625263637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,512,0.003212444484233856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,768,0.0062977779242727495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,768,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,256,0.003216000066863166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,512,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,512,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,128,0.0028453332682450614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,256,0.005984000033802456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,256,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,3072,0.009346666435400644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,128,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,32,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,128,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,768,0.003863111138343811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,65536,0.1060568888982137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,64,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,32,0.0052391112678580815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,65536,0.060933331648508705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,2560,0.009358222285906473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,16384,0.03304622239536709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,51200,0.05063644382688734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,51200,0.09984533654318915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,64,0.0028693332440323303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,12288,0.026945778065257605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,16384,0.029832889636357624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,10240,0.022755554980701868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,12288,0.01686844395266639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,12288,0.02370133333735996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,8192,0.0199946661790212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,10240,0.022835556003782485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,65536,0.06655110915501912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,8192,0.013271999855836233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,51200,0.049356444014443286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,7168,0.017708443933063082
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,1024,0.0042364444169733255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,6144,0.016216889023780823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,16384,0.020417778028382193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,7168,0.012434666355450949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,5120,0.014090667168299357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,6144,0.011424889167149862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,6144,0.014142221874660917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,4096,0.011160000330872007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,5120,0.012796444197495779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,4096,0.008711111214425828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,3584,0.009650666680600908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,3584,0.010402667025725046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,7168,0.015498666299713982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,3072,0.008720888859695857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,3584,0.008371555142932469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,5120,0.010021333065297868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,2560,0.008006222546100616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,3072,0.009365333451165093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,4096,0.012131555212868584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,2048,0.00685422205262714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,2560,0.0090844440791342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,2048,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,1536,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,10240,0.014999111493428549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,1536,0.0069448889957533936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,1024,0.003927111211750242
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,8192,0.018235555953449674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,1024,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,1024,0.00630666646692488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,768,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,2048,0.00777599960565567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,512,0.003216888962520493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,512,0.005997333261701796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,51200,0.08374489016003078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,256,0.0028684443483750024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,256,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,2560,0.007489778101444244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,256,0.005146666533417172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,768,0.003549333247873518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,64,0.0028728888266616394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,128,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,32,0.002516444358560774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,128,0.004940444396601783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,768,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,65536,0.08860888746049668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,32,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,3072,0.007734222544564142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,51200,0.0694026682111952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,65536,0.06014577547709147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,16384,0.027694221999910142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,128,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,51200,0.04751555456055535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,51200,0.04600977897644043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,16384,0.01958311100800832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,16384,0.025731555289692346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,10240,0.018569777409235638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,12288,0.016200888488027785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,12288,0.02167555524243249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,8192,0.016227554943826463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,10240,0.014469333820872836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,1536,0.006649777707126405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,8192,0.012490666574902005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,8192,0.017094223035706412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,12288,0.021634666456116572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,7168,0.014473777678277759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,64,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,6144,0.01033866653839747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,512,0.00555377784702513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,5120,0.012110222544935016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,5120,0.009373333719041612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,4096,0.010738666686746808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,7168,0.011774222056070963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,6144,0.013787555197874704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,3584,0.011370666325092316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,4096,0.008381333616044786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,4096,0.010553778045707278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,6144,0.013097777962684631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,3584,0.009000889129108852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,10240,0.02000266644689772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,65536,0.05872622463438246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,2560,0.00980088859796524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,2560,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,5120,0.01197866681549284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,2048,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,3072,0.01036088913679123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,1536,0.006274666637182236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,2048,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,3072,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,7168,0.01463644372092353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,1024,0.004884444591071871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,1536,0.006322666588756774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,768,0.004563555535342959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,1024,0.0063457778758472866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,2048,0.007633777956167857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,1024,0.006042666733264923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,512,0.004048000193304486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,768,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,768,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,512,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,512,0.0053022222386466126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,256,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,256,0.004967111266321606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,1536,0.006673777682913675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,128,0.005649777750174205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,128,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,64,0.004945777770545748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,3584,0.008033778104517195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,32,0.0028515555378463534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,65536,0.07992621925142077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,128,0.0028862222615215513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,3072,0.00868444475862715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,51200,0.06333244509167142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,65536,0.056735999054378934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,65536,0.058190219932132296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,2560,0.008350222474998897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,51200,0.0466231107711792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,51200,0.04424355427424113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,12288,0.019948444432682462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,16384,0.02532888948917389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,10240,0.017369776964187622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,12288,0.016172443826993305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,12288,0.020049777295854356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,8192,0.015236443943447538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,32,0.00508711114525795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,8192,0.012464000119103326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,7168,0.013797333670987023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,16384,0.019292443990707397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,256,0.002974222310715251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,6144,0.012486221889654795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,10240,0.014399111270904541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,6144,0.01237777786122428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,6144,0.00997866690158844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,8192,0.01578044394652049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,5120,0.00907199995385276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,4096,0.010638222098350525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,7168,0.011120888921949597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,4096,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,3584,0.010694222317801582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,10240,0.017975111802419026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,5120,0.011369778050316704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,3072,0.009119110802809397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,64,0.0025946667624844448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,3584,0.008553778131802877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,2560,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,3072,0.007691555553012424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,3072,0.007672000262472365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,2048,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,7168,0.014301333162519666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,2560,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,1536,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,4096,0.010079999764760336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,2048,0.006643555644485686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,1024,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,1536,0.006310222049554189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,3584,0.008029333419269985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,16384,0.025408888856569927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,768,0.004270222038030624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,1024,0.006516444600290722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,1024,0.005968889014588461
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,512,0.0038746665749284956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,768,0.006303999986913469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,256,0.0029519999192820657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,768,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,2560,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,128,0.002847111059559716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,256,0.0059279998143513995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,2048,0.007373332977294922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,64,0.0025351111673646504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,1536,0.006908444480763541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,32,0.0028764444092909494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,64,0.004960000101062986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,65536,0.0670657753944397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,32,0.004939555707905027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,512,0.005924444645643234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,51200,0.055108444558249585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,65536,0.05434044533305698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,256,0.004943111290534337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,128,0.005653333332803514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,128,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,51200,0.04256355431344774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,12288,0.01755288905567593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,16384,0.019135110908084445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,16384,0.02236088944805993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,10240,0.015825778245925903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,5120,0.011056889262464313
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,12288,0.019575110740131803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,8192,0.013530666629473368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,10240,0.014139556222491793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,10240,0.016698666744761996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,7168,0.012920888761679331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,8192,0.011752000285519494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,8192,0.015000000596046448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,16384,0.021640888518757288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,6144,0.011740444435013665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,51200,0.04594577683342827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,7168,0.013237333132161034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,6144,0.009715555442704095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,6144,0.011384000380833944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,12288,0.015878222054905362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,5120,0.009108444054921469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,5120,0.010610666539933948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,512,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,4096,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,3584,0.009069333473841349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,4096,0.009054222040706212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,3072,0.009240888886981541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,3584,0.007975111405054728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,3584,0.007976888782448238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,2560,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,3072,0.007351111206743453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,2560,0.007344889144102733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,2560,0.007246221933099959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,65536,0.05697688791486952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,2048,0.006867555694447623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,2048,0.007148444652557373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,7168,0.010829333629873065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,5120,0.01075111081202825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,3072,0.007380444142553542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,1536,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,1024,0.006516444600290722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,4096,0.009515555368529426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,768,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,1536,0.006268444574541516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,512,0.0042364444169733255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,512,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,1024,0.004952888935804367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,256,0.0031973332580592898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,512,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,128,0.0032097777972618737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,256,0.005688888745175467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,256,0.005247999810510212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,64,0.002696000039577484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,768,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,64,0.004958222309748332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,1536,0.006640000061856375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,65536,0.06073511309093899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,2048,0.007322666545708974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,128,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,128,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,65536,0.051968889103995435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,65536,0.05621777640448677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,16384,0.019688000281651814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,51200,0.0452497767077552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,32,0.0028391112056043413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,12288,0.01611111064751943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,16384,0.018964444597562153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,16384,0.021381333470344543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,10240,0.013539555172125498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,12288,0.018240888913472492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,768,0.004238222208287981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,8192,0.01220444424284829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,10240,0.014852444330851236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,51200,0.04760711060629951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,7168,0.011758222348160215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,8192,0.014096889230940076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,51200,0.04103822178310818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,6144,0.010471111370457543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,7168,0.010390222072601318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,7168,0.01257511145538754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,32,0.005063111169470681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,6144,0.009698666632175446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,6144,0.010417777631017896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,4096,0.008994667066468133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,1024,0.0058853332367208265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,5120,0.009017777939637503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,10240,0.013467555244763693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,3584,0.00851288851764467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,4096,0.008016889293988546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,3072,0.00793688909875022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,4096,0.008240000241332585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,8192,0.011801777614487542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,2560,0.006970666348934174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,3072,0.007070221834712558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,2560,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,2560,0.007000889215204451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,2048,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,2048,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,1536,0.005617777920431561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,2048,0.00666222224632899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,3584,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,1536,0.006912888752089606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,3584,0.007605333295133378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,1536,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,768,0.004228444563017951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,1024,0.006330666856633292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,1024,0.005643555687533484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,512,0.003915555361244413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,12288,0.015150222513410779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,768,0.005372444374693765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,256,0.0031991110493739447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,512,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,512,0.0052915554907586836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,5120,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,128,0.003014222201373842
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,256,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,64,0.002530666689078013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,128,0.0058328890138202244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,3072,0.007640000018808577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,32,0.002542222125662698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,64,0.004967999955018361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,768,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,32,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,65536,0.055439111259248525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,65536,0.049462222390704684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,51200,0.04092622134420607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,256,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,1024,0.004594666676388847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,51200,0.044871111710866295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,51200,0.03977866636382209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,12288,0.01332355539004008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,16384,0.018567111757066514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,16384,0.018571555614471436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,10240,0.011701333026091257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,12288,0.013788444300492605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,12288,0.01592177814907498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,8192,0.011168888873524137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,10240,0.01350844403107961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,16384,0.01553422212600708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,8192,0.013097777962684631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,128,0.004969777746333016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,7168,0.010393777655230628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,6144,0.009756444229020013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,5120,0.009919111099508073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,5120,0.008893333375453949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,6144,0.009705777797434065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,10240,0.012313777373896705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,4096,0.010040889183680216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,5120,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,5120,0.008346666892369589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,8192,0.01106844428512785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,7168,0.01073066641887029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,4096,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,3072,0.006274666637182236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,3584,0.007698666718271043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,3584,0.007100444700982835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,65536,0.05137155453364054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,3072,0.007322666545708974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,3072,0.0069919998447100324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,2048,0.006955555743641323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,2560,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,6144,0.009162666896979014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,2560,0.00666844430896971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,2048,0.0069848886794514125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,2048,0.00630044440428416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,4096,0.008350222474998897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,1536,0.005997333261701796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,1024,0.006445333361625671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,7168,0.010018666585286459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,2560,0.006321777900060018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,768,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,512,0.003588444242874781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,1536,0.005703111075692707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,512,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,3584,0.009354666703277165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,512,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,256,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,128,0.0028773333049482773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,768,0.003903111235962974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,128,0.005687110953860813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,64,0.0028942223224374983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,768,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,32,0.002846222163902389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,64,0.004967999955018361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,32,0.0048942222363419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,65536,0.037754668129815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,1024,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,65536,0.05509777863820394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,51200,0.03080444534619649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,65536,0.04838666650984022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,16384,0.014073777529928418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,51200,0.04450311263402303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,51200,0.038697779178619385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,12288,0.01146755533085929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,16384,0.017906667457686532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,1536,0.006664000037643645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,12288,0.013448889056841532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,10240,0.01070755554570092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,256,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,8192,0.01036977767944336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,128,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,8192,0.009860444400045607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,7168,0.009678222239017487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,8192,0.01074044406414032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,1024,0.004582222137186262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,6144,0.008886222210195329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,12288,0.014780445231331719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,10240,0.012097777591811286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,10240,0.012812444733248817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,6144,0.008674666285514832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,5120,0.008692444198661381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,4096,0.007354666789372762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,16384,0.016241777274343703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,4096,0.00832444429397583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,7168,0.010053333308961656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,7168,0.009343999955389235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,6144,0.009675555759006077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,3584,0.007288888924651676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,3072,0.006263999889294307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,5120,0.008222222328186035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,3072,0.006879111131032308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,2560,0.0057546666098965546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,3072,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,256,0.003237333355678452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,4096,0.007431999676757389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,2048,0.006275555739800136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,2560,0.006643555644485686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,2048,0.0070257774657673305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,1536,0.0052595556610160405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,2048,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,1024,0.0042364444169733255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,1536,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,1536,0.006068444500366847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,768,0.003936888857020272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,1024,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,768,0.005977777971161737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,768,0.005596444424655702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,2560,0.007366221812036302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,512,0.005928888916969299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,3584,0.007979555262459649
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,256,0.0031840000301599503
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,5120,0.007667555577225155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,128,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,256,0.00572533326016532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,256,0.00490488898422983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,1024,0.006314666734801398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,64,0.003031111011902491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,128,0.004578666554556953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,3584,0.006917333437336816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,32,0.00464088883664873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,512,0.0053182223604785065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,51200,0.022333333889643352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,65536,0.046524445215861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,51200,0.04416888952255249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,32,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,51200,0.03738933470514085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,512,0.0035573335157500375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,12288,0.012401777837011548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,65536,0.05472800135612488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,12288,0.012755555411179861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,10240,0.011408000356621213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,128,0.005591111050711737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,16384,0.015062222878138224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,8192,0.010646222366227044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,64,0.004969777746333016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,65536,0.027133334014150832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,7168,0.009975999593734741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,8192,0.009711110757456886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,7168,0.010047111246320937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,6144,0.00869333330127928
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,7168,0.009364444348547194
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,6144,0.009506666825877296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,5120,0.008031110796663497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,6144,0.008692444198661381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,10240,0.011053333679835001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,4096,0.007338666253619724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,16384,0.01589777734544542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,16384,0.016146666473812528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,5120,0.008271999657154083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,3584,0.006675555474228329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,4096,0.008330666356616551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,3072,0.006319111006127463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,3584,0.00798044436507755
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,3584,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,12288,0.01348533315791024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,3072,0.006807999892367258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,2560,0.007300444775157505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,10240,0.012079111403889127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,2048,0.0063271112740039825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,5120,0.00869422240389718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,1536,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,2048,0.00702400008837382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,8192,0.01074488874938753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,4096,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,1536,0.005965333431959152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,2560,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,1024,0.006305777778228124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,1024,0.005550222264395819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,768,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,512,0.003946666502290302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,2048,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,512,0.0052062223354975385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,1536,0.006666666517655055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,256,0.0032071111102898917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,512,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,128,0.0031902222997612427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,256,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,3072,0.007654222349325816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,768,0.003993777765168084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,64,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,32,0.0025600000388092464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,2560,0.00666133314371109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,768,65536,0.021498666869269475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,32,0.004955555415815777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,1024,0.004608889006906086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,768,51200,0.018361777067184448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,65536,0.045808000697029956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,256,0.005010666532648934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,128,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,768,16384,0.010402667025725046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,768,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,16384,0.015847999188635085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,768,12288,0.010615111225181155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,65536,0.05379111236996121
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,12288,0.012094222009181976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,768,10240,0.00964355551534229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,51200,0.036785778072145246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,128,0.004893333133724001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,768,8192,0.009356444080670675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,64,0.004864000197913912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,10240,0.011066666907734342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,768,7168,0.00832177781396442
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,8192,0.010760889285140567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,16384,0.014819555812411837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,768,6144,0.007983110845088959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,7168,0.00925066653225157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,768,5120,0.006852444261312485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,6144,0.009518221848540837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,12288,0.013442666994200813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,768,4096,0.007296000089910295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,5120,0.008991999758614434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,5120,0.008017777568764156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,768,3584,0.006960888703664144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,4096,0.007654222349325816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,3584,0.007985778152942657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,8192,0.010028444230556488
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,768,3072,0.006613333192136552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,7168,0.010046222143703038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,768,2560,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,3072,0.007640889121426477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,3072,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,768,2048,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,2560,0.007316444483068254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,2560,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,768,1536,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,2048,0.006989333364698622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,4096,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,2048,0.006237333433495627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,768,1024,0.00423555572827657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,1536,0.006609777609507243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,768,768,0.00388711111413108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,1536,0.005930666708283954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,3584,0.006980444822046492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,768,0.005994666781690385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,768,0.005330666485759947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,6144,0.008685333033402761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,512,0.005011555635266834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,51200,0.04383733206325107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,256,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,256,0.005673777725961473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,1024,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,1024,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,768,512,0.003546666767862108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,128,0.004959110998445087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,768,32,0.002554666664865282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,512,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,64,0.004829333474238713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,512,65536,0.017080888152122498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,32,0.004966222163703707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,512,51200,0.014850666125615438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,768,128,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,51200,0.03872444563441806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,51200,0.03667555583847894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,768,256,0.0028613333900769553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,16384,0.014577777849303352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,512,12288,0.008889777792824639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,16384,0.015846222639083862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,10240,0.012109333442317115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,512,10240,0.007666666474607255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,12288,0.013120888835854001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,12288,0.012129777835475074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,512,8192,0.008360889222886827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,10240,0.011824000212881299
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,128,0.00563733321097162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,10240,0.011071110765139261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,512,7168,0.008054222497675154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,8192,0.010738666686746808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,8192,0.009934222532643212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,7168,0.010060444474220276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,7168,0.009431111315886179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,512,5120,0.0069653333889113525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,6144,0.008808000220192803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,65536,0.05150577757093641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,512,4096,0.006960888703664144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,5120,0.00903911143541336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,5120,0.008006222546100616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,512,3584,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,4096,0.008153777983453538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,768,64,0.002530666689078013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,4096,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,512,3072,0.006607111129495833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,3584,0.008029333419269985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,3584,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,3072,0.007644444704055786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,3072,0.006959111326270633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,512,2048,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,2560,0.006334222025341458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,65536,0.04496444596184624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,2048,0.006276444428496891
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,512,1536,0.004230222354332606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,512,16384,0.011063999599880643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,1536,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,512,1024,0.0038844446341196695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,2560,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,1024,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,1024,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,512,768,0.0035422220826148987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,2048,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,768,0.005984000033802456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,512,512,0.0032035555276605818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,768,0.005245333330498801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,512,256,0.0031991110493739447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,1536,0.006618666566080517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,256,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,256,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,512,128,0.0028693332440323303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,128,0.00563733321097162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,512,64,0.0026773332307736077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,128,0.00490399988161193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,512,32,0.0025155554629034465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,64,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,512,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,256,65536,0.015614221493403116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,512,6144,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,256,51200,0.01310222182008955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,65536,0.04662755462858412
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,6144,0.009396444592210982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,256,16384,0.008721777962313758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,51200,0.03738933470514085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,256,12288,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,16384,0.01587288909488254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,16384,0.014527110589875115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,256,10240,0.0073831114504072405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,12288,0.012098666694429187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,10240,0.012101333174440594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,512,2560,0.006599999964237213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,65536,0.04184088773197598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,8192,0.009738666315873465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,51200,0.03396177623007033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,12288,0.01313244468635983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,7168,0.009008888569143083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,7168,0.010044444766309526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,10240,0.011000889043013254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,32,0.00462311092350218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,256,5120,0.006270222365856171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,6144,0.009587555295891231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,5120,0.00870400004916721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,5120,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,256,4096,0.005975111077229182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,4096,0.008377778033415476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,256,3584,0.005991111199061076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,4096,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,256,6144,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,256,3072,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,3584,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,3584,0.007297777467303806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,256,2560,0.0063324446479479475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,3072,0.006680888848172293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,8192,0.01073333372672399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,256,2048,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,2048,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,512,0.005274666680230035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,2048,0.006911110960774952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,256,1536,0.004563555535342959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,6144,0.008656889200210571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,256,1024,0.0038417776425679526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,1536,0.0058559998869895935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,256,7168,0.00700444479783376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,256,768,0.0035742223262786865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,1024,0.006322666588756774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,1024,0.005598222215970357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,256,512,0.0032044444233179092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,768,0.005988444305128521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,768,0.005429333282841577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,256,256,0.0028373334142896864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,512,0.005622222191757626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,256,8192,0.0069848886794514125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,512,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,256,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,3072,0.007632888853549957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,256,64,0.0026275554878844153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,128,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,128,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,2560,0.0070053330726093715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,2560,0.00665866666369968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,64,0.004895110925038655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,32,0.0047386665311124595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,65536,0.04599644409285652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,65536,0.04182666540145874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,256,128,0.0025822222232818604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,51200,0.03392444385422601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,256,0.005000000198682149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,256,32,0.0025271111064487034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,16384,0.01589777734544542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,128,65536,0.013840888937314352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,12288,0.013111111190583972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,128,10240,0.006962666908899943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,51200,0.03739022215207418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,128,16384,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,1536,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,128,8192,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,10240,0.010760000182522668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,128,12288,0.007446222007274628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,128,7168,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,8192,0.010758221977286868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,8192,0.010038221875826517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,128,6144,0.006045333213276333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,128,51200,0.012795555922720166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,7168,0.009018667042255402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,128,5120,0.006180444525347815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,6144,0.00941244430012173
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,6144,0.008344000412358178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,5120,0.008731555607583787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,5120,0.007732444339328342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,10240,0.012090666426552666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,128,3584,0.006083555519580841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,4096,0.008017777568764156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,4096,0.007583111524581909
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,128,3072,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,3584,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,3584,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,3072,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,128,4096,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,2560,0.006250666661394968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,16384,0.014470222923490735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,2048,0.006678222368160884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,2048,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,128,2560,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,1536,0.005923555543025334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,128,1024,0.003584889074166616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,1536,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,2560,0.007325333025720384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,1024,0.006303999986913469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,1024,0.005594666633341048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,768,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,768,0.005283555636803309
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,3072,0.0069520001610120135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,512,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,512,0.005949333310127258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,128,2048,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,256,0.004921777794758479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,256,0.005607111172543631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,128,128,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,12288,0.01216533366176817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,128,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,128,64,0.002570666579736604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,128,32,0.00253866674999396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,64,0.004858666823969947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,128,256,0.0028808888875775864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,64,65536,0.010404444403118556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,32,0.004997333304749595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,64,51200,0.009705777797434065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,64,16384,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,65536,0.0414631101820204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,51200,0.03363555669784546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,16384,0.014496000276671516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,128,512,0.0031928889867332247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,64,10240,0.006263999889294307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,64,8192,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,128,0.004592888885074192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,12288,0.012108444339699216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,64,7168,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,64,6144,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,10240,0.01070844464831882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,64,5120,0.005923555543025334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,8192,0.009694221946928237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,7168,0.009357333183288574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,128,768,0.0035333335399627686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,6144,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,5120,0.007989333735571967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,64,3072,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,64,12288,0.006292444550328785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,64,2560,0.006031999985376994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,128,1536,0.004575110971927643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,64,2048,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,3072,0.0069422221018208405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,64,1536,0.004270222038030624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,2560,0.006567111031876669
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,64,1024,0.003622222277853224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,64,4096,0.005777777896987067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,64,768,0.003576888806290097
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,64,3584,0.006626666833957036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,1024,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,4096,0.007352888584136963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,64,256,0.0029022221763928733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,768,0.0052728888889153796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,64,128,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,256,0.004965333475006951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,64,64,0.0026088888860411118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,128,0.004631999880075455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,64,32,0.002569777684079276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,64,0.004983110974232356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,64,512,0.0031999999450312722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,32,65536,0.009376889301670922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,32,51200,0.008279110822412703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,7168,0.010095111197895473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,32,16384,0.005970666805903117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,3584,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,51200,0.033752888441085815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,32,12288,0.006647999915811751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,2048,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,16384,0.014144000079896716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,1536,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,32,8192,0.00563733321097162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,12288,0.012131555212868584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,32,7168,0.006252444452709622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,10240,0.01071733319097095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,32,6144,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,32,0.004809777769777509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,8192,0.009717333647939894
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,32,5120,0.006010666489601135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,32,4096,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,512,0.004931555440028508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,32,3584,0.005629333357016246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,5120,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,32,3072,0.005620444400442972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,4096,0.007303111255168915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,32,10240,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,32,2560,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,3584,0.0069671107663048645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,32,2048,0.005614222337802251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,32,1536,0.004345777961942884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,3072,0.0069848886794514125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,32,1024,0.0038951109680864546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,2048,0.006218666831652324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,32,768,0.003510222252872255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,7168,0.009347555538018545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,65536,0.04180888997183906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,6144,0.008349333372380998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,32,256,0.002578666640652551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,768,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,32,128,0.002534222271707323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,512,0.0052471111218134565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,32,64,0.0025333333760499954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,256,0.004934222333961063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,32,32,0.002530666689078013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,128,0.004573333180612988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,64,0.004615111069546806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,1536,0.0057911111248864075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,1024,0.005341333233647876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,16384,0.336042669084337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,51200,0.5431457625495063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,16384,0.25716177622477215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,12288,0.25706222322252065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,2560,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,16384,0.13677422205607095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,32,0.004577777865860197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,3,32,512,0.0032302221904198327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,12288,0.1260648833380805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,10240,0.12752800517612034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,10240,0.08669066429138184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,51200,1.0159475538465712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,8192,0.0710044436984592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,8192,0.11942133638593887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,6144,0.1272053321202596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,12288,0.19280978043874106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,51200,0.4416364563835992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,7168,0.09687199857499863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,5120,0.10922577646043564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,6144,0.059784889221191406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,6144,0.08973599804772271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,7168,0.14571821689605713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,5120,0.07378222544987996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,3584,0.07300621933407254
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,4096,0.04914222160975138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,4096,0.06489155689875285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,3072,0.0633102191819085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,3584,0.045904888047112354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,3584,0.0566213329633077
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,2560,0.053223109907574125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,8192,0.17713956038157144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,3072,0.04872088962131076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,2048,0.04347733325428433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,4096,0.09028089046478271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,2560,0.03493688835038079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,2560,0.039042668210135564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,1536,0.03341689043574863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,2048,0.02884711159600152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,2048,0.03626666797531976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,1536,0.023696889479955036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,10240,0.21424267027113172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,768,0.018826666805479262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,1024,0.018222222725550335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,1024,0.02016799979739719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,512,0.013732444081041547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,768,0.015572445260153877
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,768,0.01679377754529317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,256,0.008653333617581261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,512,0.013041777743233575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,256,0.010718222293588849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,128,0.006269333263238271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,7168,0.06755377848943074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,128,0.0107351111041175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,64,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,128,0.008012444608741337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,3072,0.04070311122470432
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,512,0.01293333371480306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,64,0.008672888908121321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,5120,0.055382221937179565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,32,0.008112000094519721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,256,0.010030222435792288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,1536,0.02805333336194356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,65536,0.5120284292432997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,65536,0.968989372253418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,51200,0.7532338036431206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,51200,0.4187680085500081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,1024,0.0236799998415841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,16384,0.24937778049045137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,65536,0.4337955580817328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,12288,0.19052710798051622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,16384,0.12298666106330024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,16384,0.14993955029381648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,10240,0.15946133931477866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,12288,0.12157244152492946
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,12288,0.0838693314128452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,8192,0.13064089086320665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,10240,0.0825351079305013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,7168,0.11486577987670898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,8192,0.0921235548125373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,8192,0.0580879979663425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,6144,0.09983199834823608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,7168,0.07833866940604316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,5120,0.0903279980023702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,6144,0.04980355501174927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,4096,0.07199733124838935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,51200,0.34689778751797146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,5120,0.04523555437723795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,5120,0.05855466922124227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,3584,0.06406044297748141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,4096,0.040528890159395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,3584,0.038535111480289035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,3584,0.04304000073009067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,7168,0.05294311046600342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,3072,0.03336888882848952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,3072,0.039672000540627375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,6144,0.0816462238629659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,2048,0.03770666652255588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,32,0.005581333405441708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,2560,0.031425777408811785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,1536,0.02925866511132982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,2048,0.024452444579866197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,2048,0.030279109875361126
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,1024,0.021282666259341772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,1536,0.020051555501090158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,1536,0.023016000787417095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,768,0.015967999895413715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,1024,0.015671999918089975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,2560,0.047051555580563016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,512,0.011540444360838996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,768,0.013779555757840475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,2560,0.029330667522218492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,768,0.014114666316244336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,256,0.007298666569921706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,512,0.011312888728247749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,512,0.01145866678820716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,10240,0.10274489058388604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,256,0.009694221946928237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,64,0.0032542221662071017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,256,0.008336889247099558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,128,0.009508444203270806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,3072,0.055994669596354164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,32,0.003378666730390655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,128,0.007321777443091075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,1024,0.017267554998397827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,32,0.008008889026112026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,65536,0.31795112291971844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,51200,0.24568178918626574
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,65536,0.169852442211575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,65536,0.15195288923051622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,16384,0.08465066883299087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,12288,0.0655591090520223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,4096,0.0548017786608802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,16384,0.04126666651831733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,51200,0.12535199854109022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,10240,0.05663377708858914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,12288,0.033005333609051175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,64,0.008007110820876228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,12288,0.04659644431538052
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,10240,0.028350220786200628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,10240,0.04459377792146471
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,8192,0.024227556255128648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,8192,0.03706222110324436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,6144,0.032856888241238065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,7168,0.023364444573720295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,128,0.003879111260175705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,5120,0.028591109646691218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,7168,0.0331857767370012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,16384,0.06456710894902547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,4096,0.02439555525779724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,5120,0.02461333407296075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,8192,0.04500977860556709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,3584,0.02124622298611535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,7168,0.037767112255096436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,3584,0.018204443984561496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,3584,0.019335110982259113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,51200,0.16005067030588785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,6144,0.022037333912319605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,6144,0.029882666137483384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,3072,0.017186666528383892
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,5120,0.02106311089462704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,2560,0.015484443969196744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,2048,0.013422222601042854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,3072,0.018922666708628338
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,2048,0.014153778553009033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,1536,0.011052444577217102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,2048,0.013088000317414602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,3072,0.01699822147687276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,2560,0.016114667057991028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,1536,0.011004444625642566
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,1536,0.011942221886581846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,768,0.0070471109615431885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,1024,0.009309333231714036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,2560,0.014368888404634265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,512,0.004953777624501122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,768,0.008307555483447181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,768,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,256,0.003516444315512975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,512,0.00693866651919153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,256,0.007009777757856581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,1024,0.008685333033402761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,128,0.003216888962520493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,128,0.007002666592597961
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,64,0.002826666666401757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,4096,0.019426667028003268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,32,0.0028737777223189673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,512,0.007655999726719326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,32,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,65536,0.24032621913486055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,256,0.0059368887709246735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,65536,0.14356888665093318
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,51200,0.18722755379147002
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,65536,0.1329680018954807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,1024,0.009390222529570261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,16384,0.06641955508126153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,51200,0.16990221871270073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,12288,0.051464888784620494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,16384,0.03741066654523214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,16384,0.05708266629113091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,64,0.005935110979610019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,12288,0.035475555393430926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,12288,0.04951644606060452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,8192,0.03793155484729343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,10240,0.025568000144428674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,7168,0.0343137780825297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,8192,0.02251200046804216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,8192,0.03240622083346049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,6144,0.029836446046829224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,7168,0.027631110615200464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,51200,0.105076445473565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,5120,0.02625333269437154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,6144,0.025916443930731878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,10240,0.04391466577847799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,4096,0.022385777698622808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,5120,0.018339556124475267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,5120,0.020948444803555805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,3584,0.020390222469965618
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,4096,0.019857777489556205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,7168,0.020620443754725985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,3584,0.015650666422314115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,6144,0.019134221805466544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,3072,0.015498666299713982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,2560,0.01611022154490153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,10240,0.03770666652255588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,2560,0.013428444663683573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,2048,0.014120888378885059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,2560,0.012846222354306115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,4096,0.02162222233083513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,1536,0.00904177791542477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,2048,0.012885332935386233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,1536,0.009736888938479954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,3072,0.014482667048772177
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,1024,0.007127111156781514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,1024,0.00832266691658232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,768,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,1024,0.008374222450786168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,768,0.007639110916190677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,3584,0.01753422286775377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,512,0.003940444439649582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,2048,0.011321778098742167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,256,0.0031964443624019623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,512,0.007233777807818518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,512,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,128,0.0028400001012616684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,256,0.005674666828579373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,4096,0.016861332787407767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,128,0.006660444455014334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,128,0.005550222264395819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,32,0.00281866660548581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,128,0.005595555735958948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,768,0.007659555309348636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,64,0.0057297779454125305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,32,0.005636444522274866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,65536,0.2043386697769165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,51200,0.1604479948679606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,1536,0.010840000377760993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,65536,0.14230222172207302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,16384,0.05735377470652262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,64,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,16384,0.048447112242380776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,12288,0.04430400000678169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,16384,0.035904000202814736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,3072,0.018017777138286166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,10240,0.03812088900142246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,12288,0.028575999869240656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,12288,0.037646220790015325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,8192,0.03215466605292426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,256,0.006678222368160884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,10240,0.030240890052583482
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,7168,0.029150221082899306
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,8192,0.021003555920388963
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,8192,0.0269128895468182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,51200,0.1244479947619968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,7168,0.023240000009536743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,6144,0.01795644395881229
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,5120,0.022316444251272414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,10240,0.025016888976097107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,65536,0.12791111734178331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,4096,0.019824888971116807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,5120,0.017504000001483493
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,5120,0.01719999975628323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,3584,0.01808444493346744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,51200,0.10103288624021743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,4096,0.015296889675988091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,6144,0.02591111097070906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,3072,0.016024889217482675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,3584,0.014515555567211576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,2560,0.015122666954994202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,3584,0.014710222681363424
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,6144,0.021789333886570398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,2048,0.010984888507260216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,2560,0.0120666664507654
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,1536,0.008974221845467886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,2048,0.010424888796276517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,2048,0.011238222320874533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,1024,0.006614222294754452
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,1536,0.009344889058007134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,7168,0.019702222612169053
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,768,0.004754666652944353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,1024,0.00816000004609426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,1024,0.007359111474619971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,512,0.003544888976547453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,768,0.007626666790909237
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,768,0.007003555695215861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,512,0.00702044450574451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,512,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,256,0.006301333506902059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,1536,0.009391110804345872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,4096,0.01645777788427141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,128,0.0052844443255000645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,64,0.0028782222006056043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,2560,0.011788444386588203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,256,0.0030097777230872046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,64,0.005623111294375525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,32,0.005592888842026393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,256,0.005643555687533484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,65536,0.16872533162434897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,51200,0.12739643785688612
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,65536,0.06916622320810954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,128,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,3072,0.013442666994200813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,16384,0.04680622286266751
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,32,0.002568888788421949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,16384,0.024925332930352952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,16384,0.034747554196251765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,128,0.0028488888508743713
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,12288,0.028007109959920246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,3072,0.013110222087966071
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,10240,0.024917332662476435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,10240,0.017808000246683758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,51200,0.05674044291178385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,12288,0.03549955619706048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,7168,0.02162666618824005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,8192,0.014707555373509726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,8192,0.021033777130974665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,12288,0.018760000665982563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,7168,0.013781333135233985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,10240,0.031094223260879517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,7168,0.018280888597170513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,6144,0.013056000073750814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,6144,0.017607111069891188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,4096,0.013960889644092984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,5120,0.014398222168286642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,8192,0.025403555896547105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,51200,0.08525866932339138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,3584,0.013047999805874295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,4096,0.013459555804729462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,3072,0.01126933346192042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,6144,0.01854311095343696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,3584,0.012470222181744046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,2560,0.009739555418491364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,5120,0.01718933383623759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,3072,0.010058666268984476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,2048,0.008628444539176093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,65536,0.08907288975185818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,5120,0.012410666379663678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,2560,0.01015466699997584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,2048,0.008390222158696916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,2048,0.0096124443742964
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,4096,0.011735111474990845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,3584,0.011241777903503843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,1024,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,1536,0.007679111427730984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,768,0.003918222255176968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,1024,0.006963555183675554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,3072,0.01126933346192042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,512,0.003560888684458203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,768,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,2560,0.009005332986513773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,768,0.006301333506902059
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,256,0.0031991110493739447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,1536,0.007407111426194509
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,128,0.002887111157178879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,512,0.005821333163314396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,256,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,64,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,128,0.005964444329341252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,32,0.0025191110455327565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,128,0.004906666775544484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,1536,0.007664889097213745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,64,0.0052426668504873914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,32,0.0052471111218134565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,1024,0.00665155549844106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,51200,0.11571199364132351
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,65536,0.06548533174726698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,16384,0.04296355446179708
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,51200,0.08649688959121704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,256,0.005260444349712796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,16384,0.02184444500340356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,16384,0.0332915551132626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,10240,0.028983109527164038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,65536,0.1467368867662218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,10240,0.015947555502255756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,8192,0.024932444095611572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,65536,0.08187377452850342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,51200,0.05359199974271986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,8192,0.014025777578353882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,8192,0.02032533288002014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,7168,0.01328000012371275
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,6144,0.016473778420024447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,12288,0.026520000563727483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,6144,0.012476444244384766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,5120,0.015621332658661736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,512,0.0063075555695427795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,5120,0.011383111278216044
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,12288,0.0337119996547699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,4096,0.013080889152155982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,4096,0.010034666293197209
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,12288,0.017958222164048087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,3584,0.011193777951929303
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,4096,0.012831110921170978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,3072,0.01018311083316803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,3584,0.009521777431170145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,6144,0.01679733395576477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,2560,0.009523555636405945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,3072,0.011083555718262991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,5120,0.014352889524565803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,2048,0.008019555773999956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,7168,0.018222222725550335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,2048,0.008951110972298516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,1536,0.00665866666369968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,2048,0.007325333025720384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,3584,0.011796444654464722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,1024,0.004592888885074192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,1536,0.007341333561473423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,7168,0.020971554848882887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,1024,0.006671999891599019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,1024,0.006596444381607904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,768,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,512,0.003592888928121991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,10240,0.023544889357354905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,512,0.005618666609128316
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,3072,0.00868000007337994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,256,0.0032106666929192017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,512,0.006274666637182236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,128,0.0029102222373088202
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,256,0.005962666538026597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,768,0.00443822228246265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,128,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,128,0.004950222041871813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,32,0.002887111157178879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,64,0.005210666606823604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,65536,0.1244079934226142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,1536,0.006991110742092133
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,65536,0.07260977559619479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,64,0.0028737777223189673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,51200,0.09829955630832249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,768,0.006010666489601135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,16384,0.037696000602510236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,51200,0.07208889060550265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,32,0.0052639999323421055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,16384,0.021418665846188862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,16384,0.030211554633246526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,12288,0.01737244427204132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,65536,0.06304088566038343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,10240,0.025755556093321905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,10240,0.01551466683546702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,8192,0.02168088820245531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,51200,0.051390220721562706
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,12288,0.029959112405776978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,8192,0.01899111105336083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,7168,0.01956355571746826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,2560,0.009521777431170145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,7168,0.0127697777416971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,7168,0.016671111186345417
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,6144,0.017893332574102614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,10240,0.021708443760871887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,5120,0.01611111064751943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,6144,0.011744000017642975
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,6144,0.015255999233987598
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,4096,0.011406222151385413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,5120,0.013348444468445249
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,8192,0.013824000126785703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,4096,0.009108444054921469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,2560,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,3584,0.008904889225959778
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,3072,0.009610666996902889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,3072,0.0105253333846728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,3072,0.00834222220712238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,2560,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,3584,0.0107022225856781
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,256,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,2048,0.00756888919406467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,2560,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,2560,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,2048,0.00739555557568868
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,2048,0.008423111504978603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,1024,0.004204444587230682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,1536,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,1024,0.0063582224150498705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,768,0.0038942222793896994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,12288,0.024304888314670984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,768,0.006411555740568373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,768,0.005993777679072485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,3584,0.011081777513027191
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,512,0.005955555372767978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,512,0.005614222337802251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,256,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,1024,0.006635555376609166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,256,0.005746666755941179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,256,0.00527555536892679
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,4096,0.01221244451072481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,64,0.002586666701568498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,128,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,1536,0.006253333141406377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,64,0.004997333304749595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,1536,0.0070062221752272705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,65536,0.10672977897855972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,5120,0.010718222293588849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,51200,0.08380089203516643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,65536,0.06086933612823486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,65536,0.06649333238601685
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,16384,0.03358755509058634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,128,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,512,0.003195555466744635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,16384,0.020271110865804885
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,12288,0.026732444763183594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,12288,0.016873778568373788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,32,0.005477333234416113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,10240,0.022936888866954382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,12288,0.027831110689375136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,8192,0.020255999432669747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,10240,0.014936889211336771
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,10240,0.020288000504175823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,7168,0.01808533403608534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,8192,0.013430222041077085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,8192,0.018389332625601027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,6144,0.016388444436921012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,32,0.002537777854336633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,7168,0.012340444657537671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,5120,0.014444443914625378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,6144,0.011400000088744693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,6144,0.014850666125615438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,4096,0.01070577816830741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,5120,0.012749333348539142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,128,0.0028728888266616394
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,3584,0.009544000029563904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,51200,0.04927733209398058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,3584,0.008329778081840938
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,3584,0.01073155552148819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,5120,0.010049777726332346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,3072,0.009578666753239101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,2560,0.00797333319981893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,3072,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,4096,0.00902133352226681
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,2048,0.006855999843941794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,2560,0.007615110940403408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,51200,0.07043733199437459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,1536,0.005232000102599462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,2048,0.007325333025720384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,16384,0.02712533374627431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,2048,0.007709333466158972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,3072,0.008909333083364699
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,1536,0.006802666518423293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,768,0.003617777592606015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,1024,0.006445333361625671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,1024,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,512,0.0035422220826148987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,768,0.005941333456171884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,4096,0.012107555237081317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,512,0.005615111026499007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,256,0.0028684443483750024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,1024,0.00425511101881663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,1536,0.0069848886794514125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,128,0.0028657778683635923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,768,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,128,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,128,0.005838222387764189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,64,0.002510222295920054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,32,0.0025191110455327565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,2560,0.008053333395057255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,64,0.004935111022657818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,7168,0.015445333388116626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,32,0.004961777892377642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,51200,0.06939822435379028
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,256,0.005290666802061929
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,65536,0.05870755513509115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,16384,0.027837332752015855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,51200,0.04730400111940172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,16384,0.0270444436205758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,512,0.005982222242487802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,16384,0.01958755486541324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,65536,0.08933777941597833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,10240,0.01867288847764333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,12288,0.016163556112183463
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,12288,0.024350222614076402
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,10240,0.014516444669829475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,10240,0.01863822175396813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,7168,0.014524444937705994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,8192,0.016508445143699646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,51200,0.06622578038109674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,6144,0.013597333596812354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,7168,0.011758222348160215
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,12288,0.022316444251272414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,5120,0.012410666379663678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,6144,0.01037599974208408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,6144,0.014144889182514615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,5120,0.009015999734401703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,256,0.005976000179847081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,5120,0.012413333687517377
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,8192,0.01685244507259793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,3584,0.012574222352769641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,4096,0.008408000071843466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,4096,0.010933332973056369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,3072,0.010321777727868822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,3584,0.009120889008045197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,2560,0.009182222187519073
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,3072,0.007975111405054728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,2560,0.007355555891990662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,2560,0.007552000383536021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,2048,0.007646222081449296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,4096,0.010724444356229572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,1536,0.00592533333433999
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,65536,0.059121780925326876
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,2048,0.00702400008837382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,1024,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,8192,0.012870222330093384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,3584,0.007999999655617608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,1536,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,768,0.004562666846646203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,1024,0.006303999986913469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,3072,0.007687110867765214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,512,0.0038888889054457345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,7168,0.014474666780895658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,768,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,256,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,512,0.005958222266700532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,128,0.0028862222615215513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,256,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,256,0.0052062223354975385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,64,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,128,0.005671999934646819
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,128,0.004939555707905027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,2048,0.007391999993059371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,1536,0.006800888727108638
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,65536,0.07904533545176189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,64,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,32,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,1024,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,65536,0.05788444148169624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,65536,0.05647110939025879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,16384,0.025681777132882014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,51200,0.04657155606481764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,51200,0.060032890902625195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,16384,0.019343111250135634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,16384,0.02534488836924235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,10240,0.017533333765135873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,12288,0.02023555503951179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,10240,0.014356444279352823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,51200,0.06252533197402954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,768,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,8192,0.012443555725945367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,512,0.0053102220926019884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,12288,0.019727999965349834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,7168,0.011126221881972419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,12288,0.015998222761683993
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,6144,0.012689777546458773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,6144,0.010022222167915767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,7168,0.013745778136783175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,5120,0.011742221812407175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,32,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,5120,0.011761777930789523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,4096,0.01054133309258355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,5120,0.009097778134875828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,10240,0.018851555056042142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,4096,0.00833422193924586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,4096,0.010033778018421596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,8192,0.016568889220555622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,3584,0.007986666427718269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,3072,0.011039999624093374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,3072,0.007648888561460707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,7168,0.01423911088042789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,3072,0.007657777931955125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,2560,0.007306666837798224
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,2048,0.006959111326270633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,3584,0.011011555790901184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,2048,0.007196444604131911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,2048,0.006976889239417181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,3584,0.00906755526860555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,6144,0.013282666603724161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,1536,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,1024,0.00479644454187817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,1024,0.0063866666621632045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,768,0.004230222354332606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,1024,0.005766222046481238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,512,0.0038808890514903595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,768,0.00628977765639623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,768,0.00565244464410676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,1536,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,256,0.0031964443624019623
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,512,0.005249777601824866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,128,0.0028524444335036804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,256,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,1536,0.006653333289755716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,2560,0.009169778062237633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,64,0.0026106666773557663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,128,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,32,0.002809777855873108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,128,0.004912000149488449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,65536,0.06840533018112183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,8192,0.015153777268197803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,51200,0.054679112301932446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,65536,0.05684000253677368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,512,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,16384,0.02124533388349745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,51200,0.056848002804650195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,256,0.004966222163703707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,16384,0.019084443648656208
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,12288,0.017669333351982962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,16384,0.022467555271254644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,10240,0.015543111496501498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,12288,0.01588800052801768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,64,0.004914666629499859
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,8192,0.013818666338920593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,32,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,10240,0.014183110660976835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,65536,0.05424088901943631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,8192,0.011776000261306763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,8192,0.014179555906189812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,51200,0.04588977826966179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,7168,0.010729778144094678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,7168,0.013468444347381592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,5120,0.01074577785200543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,2560,0.007339555356237624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,5120,0.01036977767944336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,12288,0.019609777463806998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,5120,0.00903733323017756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,10240,0.016358221570650738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,3584,0.009345778160625035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,4096,0.008387555678685507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,4096,0.009307555854320526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,6144,0.011704000333944956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,3584,0.008049777812427944
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,3584,0.007999111380841997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,2560,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,3072,0.007315555380450354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,6144,0.012864888542228274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,2560,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,4096,0.010018666585286459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,7168,0.012812444733248817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,1536,0.006335111127959357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,2048,0.006624889042642381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,3072,0.00868800034125646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,1536,0.006720888945791457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,1536,0.006342222293217977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,6144,0.010041777458455827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,768,0.0047582222355736625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,1024,0.005597333527273602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,2048,0.006462222172154321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,768,0.006298666612969504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,768,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,2048,0.006959111326270633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,512,0.005936000082227919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,512,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,1024,0.004910222358173794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,256,0.005910222315125995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,128,0.0032115555885765287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,256,0.005006222261322869
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,1024,0.0063368889192740125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,128,0.005631999837027655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,64,0.0028515555378463534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,32,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,2560,0.0069848886794514125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,512,0.0038924444880750445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,65536,0.059115555551317006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,32,0.004954666727119022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,256,0.0035457776652442086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,51200,0.04790488878885905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,65536,0.0518488883972168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,16384,0.0195795562532213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,51200,0.04547022117508782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,3072,0.007709333466158972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,12288,0.01606755620903439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,16384,0.0188719994492001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,128,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,10240,0.01405066748460134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,12288,0.015562666787041558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,12288,0.01738933391041226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,64,0.004971555537647671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,8192,0.012421333127551608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,10240,0.013507555756303998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,7168,0.011425777441925473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,8192,0.01108088923825158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,8192,0.013815110756291283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,7168,0.010401777923107147
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,7168,0.01279022213485506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,5120,0.009916444619496664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,6144,0.010056888891590966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,16384,0.02164711058139801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,4096,0.009360888765917884
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,5120,0.009015111459626092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,4096,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,3584,0.008737777670224508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,65536,0.05618133147557577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,6144,0.011126221881972419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,3072,0.00832711077398724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,51200,0.050198223855760365
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,6144,0.009738666315873465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,2560,0.006657777975002925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,3072,0.007640000018808577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,3072,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,2048,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,2560,0.007295110987292395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,2560,0.006822222222884496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,1536,0.005622222191757626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,10240,0.015485333071814643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,2048,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,1024,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,1536,0.006952889263629913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,3584,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,768,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,3584,0.007324444750944774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,1024,0.005631111148330901
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,512,0.0038888889054457345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,768,0.006095110956165526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,768,0.005254222287072076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,5120,0.009199110998047723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,512,0.005128889034191768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,2048,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,128,0.0030862221287356485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,256,0.005624889085690181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,1536,0.005974222388532426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,64,0.002532444480392668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,1024,0.0063822223908371395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,32,0.002575111058023241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,64,0.004977777600288391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,256,0.003540444291300244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,65536,0.05089155501789517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,65536,0.05541777941915724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,256,0.004956444518433677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,128,0.005604444278611078
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,128,0.004908444566859139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,16384,0.01662133302953508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,51200,0.04485599862204658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,16384,0.01833066675398085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,512,0.005953777581453323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,16384,0.020564445190959506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,12288,0.013779555757840475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,4096,0.00869511150651508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,12288,0.017432888348897297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,10240,0.012472000386979846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,8192,0.011216888825098673
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,51200,0.044907556639777295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,8192,0.011060444845093621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,8192,0.012288000020715924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,7168,0.01071200023094813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,7168,0.010381333529949188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,10240,0.012338666452301873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,6144,0.009732444253232744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,6144,0.009722666607962714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,6144,0.009684444301658207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,5120,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,32,0.004905777672926585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,4096,0.010031999813185798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,5120,0.009007111191749573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,7168,0.010178666975763109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,65536,0.0494044456217024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,3584,0.0092613332801395
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,4096,0.0075804442167282104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,3584,0.00812888890504837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,5120,0.00886311133702596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,3072,0.006279110908508301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,3072,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,2560,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,3072,0.00702400008837382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,2560,0.007485333416197035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,2048,0.007063999772071838
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,2560,0.006628444625271692
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,2048,0.0069955554273393415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,1536,0.005601777798599667
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,2048,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,3584,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,1024,0.005241777747869492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,1536,0.007187555233637492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,1536,0.005962666538026597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,12288,0.013044444223244986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,768,0.004208000169859992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,1024,0.005623999983072281
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,512,0.0038951109680864546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,768,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,768,0.0052826665341854095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,512,0.005940444353553984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,512,0.005244444641802046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,256,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,10240,0.013796444568369122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,256,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,64,0.0027226667023367355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,128,0.005679111099905438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,32,0.0028595555987623003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,1024,0.006343110981914733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,64,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,32,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,256,0.0031991110493739447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,65536,0.04807644420199924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,128,0.0029191111938820947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,51200,0.04448355568779839
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,51200,0.04381066560745239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,16384,0.014112888111008538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,16384,0.0174346665541331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,65536,0.03788177834616767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,12288,0.011402666568756104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,16384,0.017863111363516915
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,10240,0.010413332945770687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,51200,0.04054311248991225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,51200,0.030209776428010728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,10240,0.012115555504957834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,8192,0.010384000009960597
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,10240,0.012795555922720166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,7168,0.009692444569534725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,8192,0.010880889164076911
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,8192,0.010077333284748925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,6144,0.009024000002278222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,7168,0.009382222261693742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,6144,0.009702222214804756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,12288,0.01346933344999949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,5120,0.007867555651399825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,12288,0.014721777704026965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,4096,0.007366221812036302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,5120,0.009020444419648912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,5120,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,3584,0.006948444578382704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,4096,0.00835111074977451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,128,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,3072,0.006617777877383762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,3584,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,3584,0.007309333317809635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,2560,0.005593777944644292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,3072,0.0069715554515520734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,4096,0.008150222400824228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,2048,0.005587555468082428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,2560,0.007505777809354994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,2560,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,1536,0.005078222188684675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,2048,0.007302222152551015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,2048,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,1024,0.004219555606444676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,4096,0.007641777396202087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,7168,0.010392888552612728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,768,0.004231111043029361
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,1024,0.006667555620272954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,1024,0.005498666730191972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,3072,0.007689778175618913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,512,0.0035422220826148987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,768,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,256,0.0028924445311228433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,768,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,512,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,65536,0.05482044484880236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,512,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,256,0.005960000058015187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,64,0.002887111157178879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,256,0.004972444640265571
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,128,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,32,0.002846222163902389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,1536,0.006828444285525217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,1536,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,64,0.004917333523432414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,32,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,65536,0.054415110084745616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,65536,0.04656889041264852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,6144,0.008775999976529015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,16384,0.014807111687130399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,51200,0.0439262224568261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,128,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,12288,0.012150222228633033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,16384,0.01584533353646596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,16384,0.015508444772826301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,10240,0.011398221883508893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,128,0.004929777648713854
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,12288,0.01276266657643848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,8192,0.010048888623714447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,10240,0.012076444096035428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,10240,0.01112711098459032
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,7168,0.009869332942697737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,8192,0.009702222214804756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,51200,0.037984887758890785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,7168,0.010057777994208867
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,12288,0.013459555804729462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,65536,0.026769777139027912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,6144,0.00867022242810991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,5120,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,6144,0.009687999884287516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,51200,0.021957332889238994
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,5120,0.00904977818330129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,5120,0.008014221986134848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,3584,0.006978666616810693
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,4096,0.008343111309740279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,4096,0.007655111451943715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,3072,0.006263111200597551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,3584,0.008051555189821456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,3584,0.0070168889231152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,2560,0.005640000104904175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,3072,0.0069955554273393415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,2048,0.006260444306664997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,2560,0.006959111326270633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,4096,0.007304000357786815
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,2048,0.006829333388143116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,8192,0.010718222293588849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,1536,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,1536,0.0058462222417195635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,1024,0.004913777940803104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,3072,0.007687999970383114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,768,0.004264889078007804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,1024,0.005611555443869696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,768,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,512,0.003531555748648114
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,2048,0.006303999986913469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,512,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,256,0.0032071111102898917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,1024,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,256,0.004934222333961063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,256,0.005620444400442972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,2560,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,1536,0.005917333480384614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,64,0.002855111120475663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,128,0.005677333308590784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,128,0.004938666605287128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,32,0.0028640000770489373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,768,65536,0.022346667117542688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,512,0.005039999882380168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,64,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,768,51200,0.019377777973810833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,65536,0.05435377690527174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,65536,0.045610666275024414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,768,16384,0.010090666512648264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,128,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,51200,0.043374223841561206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,768,12288,0.010595555934641095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,16384,0.015909334023793537
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,16384,0.014544889330863953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,768,10240,0.01017066670788659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,12288,0.013448889056841532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,12288,0.012123555772834353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,768,8192,0.00868444475862715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,10240,0.012091555529170565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,10240,0.010753778119881948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,32,0.004932444542646408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,768,7168,0.009008000294367472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,8192,0.009777777724795872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,7168,0.010030222435792288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,51200,0.03710044423739115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,768,0.0053084443012873335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,6144,0.009392889009581672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,768,5120,0.006808888994985157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,5120,0.00904622260067198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,768,4096,0.007326222128338284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,8192,0.01071733319097095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,4096,0.007700444095664554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,768,6144,0.007656888829337225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,7168,0.009371555513805812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,3584,0.008052444292439355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,3584,0.006992888947327931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,768,3072,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,5120,0.00814488861295912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,768,2560,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,4096,0.00833511104186376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,2560,0.007314666277832455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,768,2048,0.004949333353175057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,6144,0.00867555538813273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,3072,0.007682667010360294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,2048,0.006014222072230444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,3072,0.007182222273614671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,768,1536,0.004920888692140579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,2048,0.0069955554273393415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,768,1024,0.004271999829345279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,1536,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,768,3584,0.0069857777820693115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,768,768,0.003913777983850903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,1024,0.005378666851255629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,768,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,768,512,0.0034719999465677473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,512,0.0052426668504873914
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,2560,0.006654222392373615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,768,256,0.0028568889117903183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,512,0.0058986664646201665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,256,0.00565155554148886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,768,128,0.002891555635465516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,768,0.0062871111763848206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,768,64,0.0025635556214385563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,128,0.005590222362014983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,64,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,512,65536,0.016972444123691983
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,32,0.004594666676388847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,512,51200,0.015033778217103748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,65536,0.05116888880729675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,65536,0.04504266712400648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,512,16384,0.011215999722480774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,51200,0.03674666749106513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,768,32,0.0027271111806233725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,512,12288,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,1536,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,1024,0.006299555715587404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,12288,0.012119111087587146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,12288,0.013194666968451606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,256,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,10240,0.011067555182509951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,128,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,8192,0.010006222460005019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,512,7168,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,512,10240,0.007654222349325816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,7168,0.009390222529570261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,512,6144,0.007287999822033777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,51200,0.038528889417648315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,512,8192,0.008031110796663497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,512,5120,0.006679111056857639
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,6144,0.009406222237481011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,6144,0.008838222258620793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,16384,0.015850666496488783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,512,4096,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,5120,0.008985777695973715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,512,3584,0.006810666786299811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,4096,0.008336000144481659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,4096,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,3584,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,512,3072,0.0063475556671619415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,7168,0.010035555395815108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,3072,0.007696000238259633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,3072,0.006948444578382704
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,2560,0.007322666545708974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,16384,0.014513777361975776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,2560,0.006331555545330048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,5120,0.008061333662933773
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,512,1536,0.004565333326657613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,2048,0.00629066675901413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,3584,0.007339555356237624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,1536,0.006650666809744305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,512,1024,0.003906666818592283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,10240,0.011848889291286469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,512,768,0.0035457776652442086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,1024,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,1024,0.005608888963858287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,512,512,0.0032666667054096856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,512,2048,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,768,0.004942222187916438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,8192,0.010775110787815519
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,2048,0.0069351109365622205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,512,0.004958222309748332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,512,128,0.0028604444944196274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,256,0.005617777920431561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,256,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,1536,0.005591111050711737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,128,0.00563733321097162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,512,2560,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,512,32,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,256,65536,0.01586666703224182
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,64,0.004924444274769889
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,768,0.006028444402747684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,32,0.0047457776963710785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,256,51200,0.013421333498424955
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,65536,0.046471112304263644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,512,256,0.002871111035346985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,256,16384,0.008632000121805403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,512,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,256,12288,0.008298666940795051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,51200,0.034304890367719866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,16384,0.014134221606784396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,512,64,0.0026657777941889236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,256,10240,0.007651555869314406
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,128,0.004909333255555895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,10240,0.012083555261294046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,10240,0.011114666859308878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,256,8192,0.007152888509962294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,65536,0.04219555523660448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,8192,0.010762666662534079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,256,7168,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,8192,0.010056888891590966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,16384,0.015848888291252982
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,7168,0.010083555347389644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,12288,0.013175110850069257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,12288,0.012069332930776807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,6144,0.009495110975371467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,6144,0.008678221868144141
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,256,4096,0.005933333188295364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,5120,0.008798222574922774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,5120,0.008064000142945183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,256,3584,0.006462222172154321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,4096,0.007630222373538547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,51200,0.03737688726849026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,3584,0.007984889050324758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,256,6144,0.006657777975002925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,7168,0.009346666435400644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,256,2560,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,2560,0.006632888896597757
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,4096,0.008249777886602614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,2560,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,256,3072,0.005970666805903117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,3584,0.007234666910436418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,2048,0.006919999917348226
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,2048,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,3072,0.007642666498819987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,256,1024,0.0038746665749284956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,1536,0.0063279999627007385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,1536,0.005966222120655908
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,1024,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,1024,0.005487111293607288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,256,512,0.0032382222513357797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,768,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,256,2048,0.005620444400442972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,256,256,0.002874666617976295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,512,0.0052844443255000645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,256,1536,0.004583999928500918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,256,128,0.0026008888251251644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,256,0.0058453331391016645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,256,0.004954666727119022
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,128,0.005680888891220093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,256,768,0.0035253332720862497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,128,0.00489688871635331
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,768,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,64,0.004966222163703707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,128,65536,0.014496889379289416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,32,0.004960888789759742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,128,51200,0.012443555725945367
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,512,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,65536,0.041802667909198336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,128,16384,0.007791111038790808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,51200,0.033901333808898926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,256,5120,0.006133333262470033
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,256,64,0.0025155554629034465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,256,32,0.0025884444928831523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,12288,0.013140444126394061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,128,10240,0.00665244460105896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,12288,0.012220444778601328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,65536,0.04589777853753832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,128,8192,0.006739555547634761
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,10240,0.01090400003724628
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,51200,0.037343111303117536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,128,7168,0.006508444332414203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,8192,0.010498666928874122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,8192,0.00996088898844189
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,16384,0.014521777629852295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,7168,0.009368889033794403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,6144,0.00941955546538035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,128,5120,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,5120,0.00814577771557702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,3072,0.006804444309737947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,128,4096,0.0057706667317284485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,128,12288,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,128,6144,0.006271111054552927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,128,3584,0.006296000132958095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,4096,0.007880888879299164
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,4096,0.007606222397751278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,128,3072,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,6144,0.00869422240389718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,3584,0.007313778003056844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,128,2560,0.006284444282452266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,3072,0.007639110916190677
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,3072,0.006718222051858902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,128,2048,0.005934222290913264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,2560,0.006331555545330048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,16384,0.015863110621770222
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,7168,0.010050666828950247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,2048,0.0069546666410234236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,1536,0.006640888750553131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,3584,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,1024,0.0052577778697013855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,10240,0.012223111258612739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,1024,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,5120,0.00867022242810991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,128,1536,0.004602666530344221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,2048,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,128,512,0.003185777821474605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,768,0.00600888869828648
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,768,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,128,1024,0.003584889074166616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,512,0.005977777971161737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,512,0.004931555440028508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,128,128,0.0028560000161329904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,256,0.004982222285535601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,2560,0.007370666497283512
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,128,64,0.002565333412753211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,128,0.005619555711746216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,128,32,0.0025271111064487034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,64,0.004945777770545748
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,64,65536,0.010447111394670276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,32,0.00490044429898262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,1536,0.005728000154097875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,64,51200,0.009682666924264696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,64,16384,0.0069573331210348345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,256,0.00563822231358952
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,65536,0.041801777150895864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,64,12288,0.006189333481921091
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,51200,0.033399111694759794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,16384,0.014504000544548035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,128,0.004940444396601783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,64,8192,0.006984000404675801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,64,7168,0.00628355559375551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,8192,0.009703111317422655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,64,6144,0.0063271112740039825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,128,256,0.0029084444459941653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,64,5120,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,6144,0.008458666503429413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,64,10240,0.005988444305128521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,64,4096,0.005617777920431561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,5120,0.007987555530336168
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,64,3072,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,128,768,0.003560888684458203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,4096,0.007330666813585494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,10240,0.01072355525361167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,64,2048,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,3072,0.006622222148709827
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,64,1536,0.004557333472702238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,2560,0.006462222172154321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,7168,0.009348444640636444
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,64,1024,0.003863111138343811
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,64,768,0.0035439998739295532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,1024,0.005311111195219888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,64,3584,0.005994666781690385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,768,0.005240889059172736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,64,256,0.002854222224818336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,64,2560,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,64,128,0.002528888897763358
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,512,0.0053075556125905775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,64,64,0.002569777684079276
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,256,0.004930666751331753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,64,32,0.002559111143151919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,128,0.004927999857399198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,32,65536,0.009020444419648912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,64,0.004962666581074397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,32,51200,0.008384000096056197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,32,0.004898666507667965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,1536,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,32,16384,0.006816888848940532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,65536,0.041287110911475286
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,12288,0.012104888757069906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,51200,0.03370489014519586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,16384,0.014311111635631986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,32,8192,0.005819555371999741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,10240,0.010793777803579966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,32,7168,0.005630222045713001
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,8192,0.009528000321653154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,32,6144,0.006625777731339137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,7168,0.009351111120647853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,32,12288,0.006623111251327727
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,6144,0.008349333372380998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,32,4096,0.005619555711746216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,3584,0.006990222467316522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,32,3584,0.005635555419656966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,5120,0.007992888490358988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,32,3072,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,4096,0.007317333585686154
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,32,2560,0.005956444475385878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,3584,0.0069857777820693115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,32,2048,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,2560,0.006287999865081575
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,32,1536,0.004564444637960858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,2048,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,32,10240,0.006339555399285422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,12288,0.011752888560295105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,32,768,0.003256888853179084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,1024,0.005344888816277186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,2048,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,768,0.0052177777720822235
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,3072,0.006955555743641323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,32,256,0.0026071110947264563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,32,128,0.0026453334010309647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,64,512,0.003185777821474605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,32,5120,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,32,64,0.002551111082235972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,256,0.004926222066084544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,32,1024,0.003912888881233004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,32,32,0.0028364445186323593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,1536,0.005584888988071018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,64,0.004622222234805425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,32,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,2,32,512,0.003240888938307762
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,16384,0.3363528781467014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,51200,1.0133778254191081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,128,0.004607111215591431
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,12288,0.25603199005126953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,16384,0.1565608845816718
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,51200,0.4255679978264703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,512,0.004931555440028508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,12288,0.12404089503818089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,10240,0.10735111104117499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,8192,0.1752844386630588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,10240,0.2150933345158895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,8192,0.09061066971884833
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,7168,0.1454355584250556
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,6144,0.1261475616031223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,7168,0.08349333206812541
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,6144,0.07271377907858954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,5120,0.10756266779369777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,16384,0.1791564491060045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,5120,0.06263377931382921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,4096,0.08953866693708633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,12288,0.1876577801174588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,51200,0.5471599896748861
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,3584,0.07270310984717475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,4096,0.06677777899636163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,3584,0.05804444683922661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,3072,0.06325244241290622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,8192,0.11733688248528375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,4096,0.05132977830039131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,2560,0.053322666221194796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,10240,0.12992444303300646
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,2560,0.04259555538495382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,2560,0.034061332543691
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,2048,0.04353422257635328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,3072,0.05046755406591627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,1536,0.03427022364404466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,5120,0.0764755540423923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,3072,0.039692444933785334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,1024,0.023891554938422307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,1536,0.023231999741660223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,1536,0.02721422248416477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,3584,0.04611733224656847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,1024,0.017720000611411203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,1024,0.019991111424234178
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,512,0.013434666726324292
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,768,0.015077332655588785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,768,0.01688888834582435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,256,0.008692444198661381
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,512,0.012989333106411828
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,512,0.012585777375433179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,2048,0.03635288940535651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,256,0.01072177787621816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,256,0.009697777529557547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,64,0.0059057776298787855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,128,0.0107004443804423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,32,0.005577777822812398
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,128,0.007974222302436829
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,64,0.008029333419269985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,6144,0.0864844454659356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,32,0.007996444072988298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,2048,0.029317332638634577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,128,0.006267555471923616
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,65536,0.9679857889811198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,768,0.0192239996459749
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,51200,0.4087173408932156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,16384,0.1442355579800076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,65536,0.43538133303324383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,51200,0.37983910242716473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,7168,0.10110933250851101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,12288,0.12153600321875678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,12288,0.10157689121034409
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,16384,0.24839289983113608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,8192,0.1299848953882853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,10240,0.10189777612686157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,16384,0.12800178262922499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,12288,0.1898933384153578
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,8192,0.07492799891365899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,51200,0.7518302069769965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,7168,0.06924711333380805
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,10240,0.1590835518307156
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,6144,0.09943999846776326
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,10240,0.0856284432941013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,5120,0.08975644244088067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,7168,0.11384622255961101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,5120,0.04952355557017856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,4096,0.07167733377880521
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,65536,0.545255978902181
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,3584,0.06369333134757148
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,4096,0.05434933304786682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,7168,0.07909866836335924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,3584,0.036644445525275335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,6144,0.06997155480914645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,3072,0.04172711239920723
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,3072,0.032334221733940974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,8192,0.09209244118796454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,2560,0.04656711220741272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,4096,0.04236355423927307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,2048,0.03762755460209317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,2560,0.027899556689792212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,6144,0.05964799722035726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,3072,0.05632266733381483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,1536,0.029255112012227375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,2048,0.024140444066789415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,2048,0.030230220821168687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,1536,0.01990133358372582
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,5120,0.06127022372351753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,1536,0.022628444764349196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,1024,0.015525332755512662
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,768,0.013519110778967539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,768,0.013816888961527081
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,2560,0.035411556561787925
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,512,0.011048888994587792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,512,0.011115555961926779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,768,0.01564088960488637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,128,0.0038844446341196695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,1024,0.01718488832314809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,512,0.011391999820868174
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,64,0.0031991110493739447
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,128,0.00942488925324546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,128,0.00701422244310379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,32,0.0032604444358083936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,1024,0.020941333638297185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,64,0.007341333561473423
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,32,0.007495111061467065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,51200,0.24709865781995985
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,256,0.009409777820110321
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,65536,0.1535200013054742
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,16384,0.0846666693687439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,3584,0.048017776674694486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,51200,0.18995111518436006
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,256,0.007288888924651676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,65536,0.3184044361114502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,12288,0.0648497806655036
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,10240,0.05706666575537788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,256,0.008020444048775567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,12288,0.041512889994515315
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,8192,0.04500977860556709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,10240,0.036433776219685875
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,10240,0.04352710975541008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,7168,0.03812355465359158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,51200,0.1353706651263767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,8192,0.03135644396146139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,8192,0.035007112556033664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,6144,0.03288355469703674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,16384,0.050195554892222084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,7168,0.032837334606382586
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,6144,0.026208889153268602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,6144,0.027678221464157104
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,12288,0.05059911145104302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,4096,0.024656888511445787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,65536,0.2203528881072998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,3584,0.021272889441914026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,5120,0.024774221910370722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,4096,0.021909332937664453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,3072,0.018756444255510967
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,3584,0.020071999894248117
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,3072,0.016691555579503376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,7168,0.030396444929970637
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,5120,0.02854577700297038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,5120,0.023152889476882085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,2560,0.015508444772826301
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,2048,0.01368799971209632
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,4096,0.02040799955526988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,1536,0.011343110766675739
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,2048,0.013295110729005603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,3584,0.018837332725524902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,2048,0.013639999760521783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,1536,0.011247999966144562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,1536,0.011623111036088733
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,768,0.007299555672539606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,1024,0.009404444032245213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,1024,0.009353777600659264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,512,0.004905777672926585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,768,0.008394666843944127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,768,0.007376000285148621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,256,0.0035617777870761026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,512,0.006565333240562015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,256,0.006807999892367258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,1024,0.008323555191357931
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,128,0.0032248888164758682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,128,0.006783999916579988
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,16384,0.07054222292370267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,128,0.005255110975768831
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,32,0.0028417776856157514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,2560,0.015152888165579902
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,64,0.005463111317820019
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,32,0.005587555468082428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,65536,0.23939556545681426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,256,0.005568888866239124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,51200,0.1861333317226834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,65536,0.13384000460306802
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,3072,0.017645332548353408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,16384,0.0657679968410068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,51200,0.13554755846659342
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,51200,0.10690666569603814
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,512,0.007634667058785756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,16384,0.05805155303743151
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,12288,0.0344177782535553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,12288,0.042825778325398765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,65536,0.18252710501352945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,2560,0.015831111205948722
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,10240,0.029263112280103896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,8192,0.039263112677468195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,8192,0.025210666987631056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,16384,0.04448799954520332
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,7168,0.03467999895413717
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,7168,0.02437066700723436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,7168,0.027004443936877783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,10240,0.04314400090111626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,6144,0.022611555125978258
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,6144,0.024244444237815008
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,10240,0.036660446061028376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,5120,0.01965688831276364
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,5120,0.021871111459202234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,6144,0.030681778987248737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,4096,0.020037333170572918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,4096,0.016963554753197562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,64,0.0028684443483750024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,5120,0.02630311085118188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,3584,0.01569955547650655
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,3584,0.017842666970358956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,2560,0.016294222739007738
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,8192,0.032122668292787336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,2560,0.012794666820102267
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,2048,0.014196443888876172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,3584,0.020397333635224234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,2048,0.012784889174832238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,2048,0.011691555380821228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,3072,0.014475555883513557
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,12288,0.05156622330347697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,1536,0.010724444356229572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,1024,0.007310222420427535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,2560,0.014411555396185981
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,768,0.005931555396980709
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,1024,0.008461332983440822
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,1024,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,512,0.0038933331767717996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,768,0.008002666963471307
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,768,0.006976000136799282
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,256,0.0032231110251612137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,512,0.006299555715587404
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,4096,0.022942221826977197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,128,0.0031786666562159858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,256,0.005372444374693765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,128,0.006648889018429651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,64,0.0028568889117903183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,128,0.0052915554907586836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,32,0.002926222152180142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,3072,0.01819644371668498
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,1536,0.009022222624884712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,64,0.005274666680230035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,32,0.0052986666560173035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,3072,0.01643288963370853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,51200,0.15770399570465088
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,1536,0.00981600003110038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,16384,0.056735999054378934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,51200,0.10700444380442302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,16384,0.047592888275782265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,12288,0.044008000029457935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,512,0.007375111182530721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,65536,0.20246133539411756
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,10240,0.03831288880772061
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,12288,0.03696444299485948
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,65536,0.15885778268178305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,8192,0.03170844581392076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,10240,0.028357333607143823
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,10240,0.03119200136926439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,51200,0.13705066839853922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,7168,0.02917511264483134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,8192,0.02416977783044179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,8192,0.02660177813635932
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,7168,0.022936888866954382
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,16384,0.04072533382309808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,7168,0.02369333306948344
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,5120,0.02269422180122799
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,6144,0.02052888936466641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,6144,0.020503999458418954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,5120,0.018389332625601027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,5120,0.018444443742434185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,3584,0.018251554833518136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,4096,0.01668444441424476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,3584,0.014517333772447375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,6144,0.02650044361750285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,3072,0.013788444300492605
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,4096,0.020248888267411124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,12288,0.03256888853179084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,2560,0.01549600064754486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,65536,0.1256542205810547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,2048,0.011094222466150919
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,2560,0.011517333487669626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,2560,0.012627555264366997
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,1536,0.009289777941173976
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,3584,0.015270221564504834
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,3072,0.016918222109476726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,1536,0.009023110899660323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,1536,0.009316444396972656
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,256,0.00675999994079272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,3072,0.013172444370057849
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,768,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,1024,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,768,0.007327111230956183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,512,0.003535110917356279
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,512,0.005943999936183293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,1024,0.00639111093348927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,256,0.003220444545149803
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,512,0.006968888971540663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,4096,0.015974221958054435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,1024,0.007636444436179266
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,256,0.0052844443255000645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,2048,0.01038755559258991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,64,0.0028417776856157514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,128,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,128,0.005247999810510212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,32,0.0028719999310043124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,65536,0.1659191052118937
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,32,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,256,0.006631111105283101
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,65536,0.11503556039598252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,51200,0.12714132997724745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,768,0.00665511108107037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,16384,0.04663377669122484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,51200,0.09343644645478989
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,12288,0.034930666287740074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,16384,0.0251511103577084
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,64,0.0052897776994440295
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,10240,0.03126044405831231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,12288,0.021391999390390184
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,12288,0.031890667147106595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,10240,0.018937778141763475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,10240,0.02370133333735996
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,2048,0.01108444482088089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,8192,0.021015110943052504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,51200,0.05995733208126492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,16384,0.035740445057551064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,7168,0.01828177769978841
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,8192,0.025460445218616064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,6144,0.01649688846535153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,5120,0.01719377769364251
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,6144,0.014813333749771118
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,8192,0.016364443633291457
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,4096,0.014523555835088095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,5120,0.01347555551264021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,5120,0.014534221755133735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,3584,0.013187555803192986
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,4096,0.012215110990736218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,4096,0.01293244461218516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,3072,0.011974222130245633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,3584,0.011222222612963783
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,3584,0.012473777764373355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,128,0.0028880000528362063
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,2560,0.010337777435779572
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,2560,0.009194666312800514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,2048,0.008967999782827165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,2048,0.009377777576446533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,6144,0.019354666272799175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,1536,0.007234666910436418
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,2048,0.008700444466537898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,65536,0.07191822263929579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,1024,0.005588444570700328
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,1536,0.007827555967701806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,3072,0.011014222270912595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,768,0.004576000074545543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,1024,0.0064399999876817065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,768,0.006610666712125142
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,768,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,512,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,512,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,256,0.0032497776879204642
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,512,0.0052373334765434265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,1536,0.007631999750932057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,256,0.005954666684071223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,256,0.004944889081848992
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,2560,0.010465777582592435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,64,0.0029004443850782183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,128,0.005976888868543837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,128,0.004576888763242298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,32,0.0028577778074476454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,7168,0.016136000553766888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,64,0.004979555391603046
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,32,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,3072,0.010046222143703038
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,65536,0.10293955273098415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,65536,0.06665511263741387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,128,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,1024,0.007071110937330458
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,7168,0.022135110365019903
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,51200,0.08596711026297675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,16384,0.021811554829279583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,12288,0.01793244481086731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,65536,0.14508977201249865
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,10240,0.028897778855429754
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,51200,0.11429511176215278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,10240,0.016027554869651794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,8192,0.025096888343493145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,16384,0.04244000050756666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,51200,0.05316266748640272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,8192,0.014663999279340109
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,12288,0.03343644407060411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,8192,0.020111110475328233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,16384,0.03358577688535055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,6144,0.018607111440764535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,7168,0.014152889450391134
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,5120,0.014485332700941296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,6144,0.01293244461218516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,6144,0.016177778442700703
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,4096,0.013531555732091268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,5120,0.01346666696998808
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,10240,0.022934221559100684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,3584,0.01201600001917945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,4096,0.010633777413103314
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,7168,0.019747555255889893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,3072,0.01037599974208408
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,3584,0.011854222251309289
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,7168,0.01792622274822659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,12288,0.03011200163099501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,2560,0.009813333551088968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,5120,0.011784888803958893
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,2560,0.008347555167145198
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,2048,0.008364444805516137
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,2560,0.010043555663691627
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,1536,0.007301333049933116
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,2048,0.007818666597207388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,4096,0.01276622215906779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,2048,0.008358221915033128
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,1536,0.006966222491529252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,1536,0.007320889168315464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,768,0.004192000048028098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,1024,0.0063279999627007385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,3072,0.008646222452322641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,512,0.0036115555299652945
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,768,0.006311999840868845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,768,0.005944889038801193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,512,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,512,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,128,0.003019555575317807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,256,0.005076444397370021
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,1024,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,64,0.0028479999552170434
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,128,0.005938666562239329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,32,0.002887111157178879
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,1024,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,64,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,65536,0.1238666640387641
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,32,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,256,0.0032213332338465583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,51200,0.10028711292478774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,65536,0.06331555710898505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,65536,0.09820622205734253
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,256,0.005947555518812603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,16384,0.038375112745496966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,51200,0.05128622386190626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,16384,0.021359110871950786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,3584,0.009588444398509132
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,16384,0.031530665026770696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,12288,0.017646221650971305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,12288,0.02844088938501146
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,8192,0.022006221943431433
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,10240,0.020632000433074105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,8192,0.013783111340469785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,7168,0.01979111135005951
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,8192,0.018930666976504855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,51200,0.07220977544784546
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,6144,0.017462222112549674
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,7168,0.01683111157682207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,128,0.004594666676388847
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,5120,0.015840000576443143
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,6144,0.012074666718641916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,6144,0.015207111835479736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,5120,0.01073333372672399
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,5120,0.013153778182135688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,4096,0.009123555488056606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,4096,0.0124177775449223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,12288,0.02962488929430644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,10240,0.025823111335436504
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,3584,0.00870666652917862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,3072,0.009347555538018545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,10240,0.015847999188635085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,4096,0.011717333561844297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,3072,0.010666666759385003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,2560,0.007822222179836696
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,2048,0.007553777760929531
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,2560,0.009376000199053023
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,7168,0.01275466630856196
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,1536,0.006611555400821898
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,3584,0.0114720000161065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,1536,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,3072,0.010760000182522668
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,2560,0.00866844422287411
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,1024,0.006619555668698416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,768,0.00387999994887246
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,3584,0.010431999961535135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,768,0.0062773335311147906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,2048,0.0075342224703894714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,768,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,512,0.003573333223660787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,2048,0.008063111040327284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,256,0.0032346666687064697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,512,0.005961777849329843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,1536,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,3072,0.008832889298597971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,256,0.005979555762476391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,256,0.004982222285535601
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,64,0.0025333333760499954
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,128,0.005927111125654644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,1024,0.005960888746711943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,128,0.004578666554556953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,65536,0.1056355569097731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,64,0.004983110974232356
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,32,0.004592000196377437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,51200,0.08398222261004978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,65536,0.06091377470228407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,65536,0.077674667040507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,16384,0.03269244564904107
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,51200,0.04928177926275465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,51200,0.07005866368611653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,12288,0.026375111606385972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,32,0.0025235555238193935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,16384,0.02036533256371816
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,10240,0.02272266646226247
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,12288,0.016918222109476726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,8192,0.019935111204783123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,12288,0.023644444015291
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,10240,0.019262222780121695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,8192,0.013167111410035027
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,8192,0.017480888300471835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,1024,0.004578666554556953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,6144,0.01646222174167633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,7168,0.012510221865442065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,7168,0.016467556357383728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,5120,0.01420177850458357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,16384,0.027108444107903376
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,6144,0.013806222213639153
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,4096,0.011062222222487131
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,5120,0.012774222426944308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,10240,0.015047111445003085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,3584,0.009351111120647853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,4096,0.008984000318580205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,4096,0.011780444118711682
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,3072,0.008808888494968414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,3584,0.008003555238246918
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,3584,0.011008000208271874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,3072,0.009595555563767752
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,2560,0.007814221911960179
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,6144,0.011448000040319232
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,2048,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,2560,0.007633777956167857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,2560,0.008341333104504479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,1536,0.005299555758635203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,2048,0.007350222104125553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,7168,0.0181040002240075
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,1024,0.0041644444896115195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,512,0.0052560000783867305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,3072,0.007702222300900354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,768,0.003914666672547658
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,1024,0.006000888844331105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,768,0.006280888699822956
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,5120,0.010040000081062317
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,512,0.0035457776652442086
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,2048,0.007319110963079665
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,256,0.0032302221904198327
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,512,0.005984000033802456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,512,0.005271111097600725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,1536,0.00701155596309238
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,128,0.00286666676402092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,256,0.004937777916590373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,1024,0.006578666468461354
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,128,0.0047111109726958806
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,32,0.0025253333151340485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,64,0.004936888813972473
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,768,0.005609777652555042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,65536,0.08913422293133205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,32,0.004942222187916438
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,51200,0.06863555643293592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,128,0.0028568889117903183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,65536,0.08177333407931857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,1536,0.006697777658700943
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,16384,0.0278479986720615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,64,0.0028435554769304064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,51200,0.059614221254984535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,12288,0.022335999541812476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,16384,0.019612444771660697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,16384,0.025824889540672302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,10240,0.01962222158908844
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,12288,0.016182222300105624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,8192,0.016569778323173523
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,10240,0.014483556151390076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,10240,0.018237334158685472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,7168,0.014471999473041959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,8192,0.017532444662517972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,65536,0.05867111020618015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,6144,0.013440888788965015
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,7168,0.01516355574131012
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,128,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,6144,0.01036444471942054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,5120,0.012158222496509552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,12288,0.02156266735659705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,5120,0.00906133320596483
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,5120,0.011911110745535957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,256,0.005924444645643234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,7168,0.011736888852384357
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,51200,0.04758666621314155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,4096,0.01111288865407308
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,4096,0.008387555678685507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,3072,0.01109688894616233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,3584,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,3584,0.009349333743254343
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,3072,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,3072,0.008731555607583787
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,8192,0.012768000364303589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,4096,0.010799110763602786
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,2048,0.00814933329820633
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,2560,0.008007999923494127
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,1536,0.006275555739800136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,2048,0.007346666521496243
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,6144,0.012996444271670448
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,1536,0.006621333460013072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,1024,0.0052986666560173035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,2560,0.009346666435400644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,768,0.004567111117972268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,1024,0.005913777897755305
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,512,0.003924444317817688
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,2560,0.00758755538198683
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,768,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,256,0.0032222221295038858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,2048,0.0069955554273393415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,512,0.005261333452330695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,256,0.005959110955397288
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,256,0.004920000003443824
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,128,0.005932444499598608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,128,0.004597333156400257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,1024,0.006404444575309753
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,32,0.002558222247494592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,64,0.004601777841647466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,768,0.005312888986534542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,512,0.006008000009589725
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,3584,0.012757333616415659
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,128,0.0028648889727062653
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,51200,0.06226133637958103
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,1536,0.006691555596060223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,65536,0.06635822190178765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,16384,0.025408888856569927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,51200,0.04667466547754076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,16384,0.019354666272799175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,16384,0.024309333827760484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,12288,0.01591377788119846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,12288,0.020768000019921195
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,65536,0.07966666751437716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,10240,0.014191110928853353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,65536,0.05792355537414551
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,8192,0.015491555134455362
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,8192,0.012372444073359171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,64,0.002546666603949335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,7168,0.013823111024167804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,8192,0.0162782222032547
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,6144,0.012432888978057437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,7168,0.011083555718262991
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,7168,0.014175999495718213
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,32,0.004905777672926585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,6144,0.013154666456911298
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,5120,0.009114666945404476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,5120,0.010821333361996545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,4096,0.010254222485754225
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,51200,0.06329244375228882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,4096,0.008616888688670265
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,3584,0.012135999898115793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,4096,0.009279111193286048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,3072,0.010055999788973067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,3584,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,3584,0.009175111022260454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,6144,0.009719111025333405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,3072,0.007662222617202335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,3072,0.007701333198282454
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,2048,0.007627555893527136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,2560,0.007322666545708974
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,2560,0.007340444458855524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,1536,0.006075555665625467
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,2048,0.006639110959238476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,12288,0.019949333535300363
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,1024,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,1536,0.006671111202902264
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,1536,0.005965333431959152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,768,0.004586666822433472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,1024,0.006324444380071428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,10240,0.017469333277808297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,1024,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,512,0.0039022221333450745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,768,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,256,0.0032088889016045462
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,10240,0.017851556340853374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,2048,0.007063111497296228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,128,0.002732444347606765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,256,0.005963555640644497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,256,0.004838222016890843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,128,0.005686222265164058
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,128,0.004648000001907349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,5120,0.011314666933483548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,32,0.002862222285734283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,64,0.004574222283230888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,512,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,512,0.004925333377387789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,32,0.004744889007674323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,51200,0.05556622478697035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,65536,0.05699377589755588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,64,0.002530666689078013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,51200,0.0531582236289978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,2560,0.009712888962692684
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,768,0.006311999840868845
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,12288,0.01754577789041731
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,16384,0.01907199952337477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,65536,0.06804088751475017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,10240,0.015820443630218506
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,12288,0.01588800052801768
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,12288,0.018869333797030978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,8192,0.013544888959990608
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,10240,0.01401600076092614
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,10240,0.015303111738628812
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,7168,0.01276088919904497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,51200,0.04609422220124138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,8192,0.014584889014561972
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,6144,0.011752888560295105
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,7168,0.010602666272057427
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,7168,0.012709333664841123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,5120,0.01075644459989336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,6144,0.009684444301658207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,6144,0.011747555600272285
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,4096,0.009741333623727163
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,5120,0.009972444838947719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,65536,0.0634595553080241
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,3584,0.009369778136412302
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,4096,0.00868266655339135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,4096,0.008832000195980072
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,3072,0.009672889278994666
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,16384,0.02176711128817664
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,2560,0.007983999947706858
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,3584,0.00795999997191959
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,3072,0.007658667034573025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,2048,0.007281777759393056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,2560,0.007318221860461765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,16384,0.02405066622628106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,2560,0.007116444408893585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,2048,0.007044444481531779
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,5120,0.008999110923873054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,2048,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,1024,0.005576888720194499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,1536,0.006623999940024481
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,768,0.004908444566859139
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,1024,0.0064088888466358185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,1024,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,512,0.003938666648334927
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,768,0.0052737775776121355
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,3072,0.007640000018808577
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,512,0.005971555494599872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,512,0.005099555684460534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,1536,0.006286222073766921
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,128,0.003195555466744635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,256,0.004605333424276776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,1536,0.005985777825117111
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,128,0.005937777873542573
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,128,0.004641777939266629
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,8192,0.011975111232863532
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,768,0.006305777778228124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,65536,0.05987733602523804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,64,0.004571555389298333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,32,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,256,0.0032355555643637977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,51200,0.046794666184319385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,65536,0.05604622099134657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,65536,0.0605306691593594
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,256,0.005951111101441913
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,51200,0.04525599877039591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,51200,0.05120799938837687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,64,0.0028799999919202593
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,32,0.002529777793420686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,3584,0.008378666308191087
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,12288,0.01610133383009169
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,10240,0.013380444712109037
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,8192,0.012418666647540199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,10240,0.014730667074521383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,16384,0.019564444820086162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,7168,0.01127911110719045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,8192,0.01109866632355584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,8192,0.012768000364303589
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,16384,0.01889155473974016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,7168,0.010377777947319878
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,16384,0.02130488885773553
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,7168,0.011808888779746162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,6144,0.009749333063761393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,6144,0.010395555860466426
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,4096,0.008995555341243744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,5120,0.00905777762333552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,5120,0.008389333056079017
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,3584,0.008567111359702216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,4096,0.008377778033415476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,4096,0.007683555285135905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,3584,0.008258666429254744
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,12288,0.01534844438234965
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,3584,0.007522666619883643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,5120,0.010083555347389644
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,3072,0.007705777883529663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,2560,0.007338666253619724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,12288,0.015134221977657743
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,2048,0.006272888845867581
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,2560,0.007477333148320516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,2560,0.006687110910813014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,1536,0.006653333289755716
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,2048,0.006291555447710886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,1536,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,6144,0.010692444112565784
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,10240,0.013807999591032663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,3072,0.0075351107451650835
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,768,0.004574222283230888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,1024,0.006647111227114995
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,2048,0.006958222223652734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,512,0.003879111260175705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,1536,0.005635555419656966
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,512,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,256,0.0033733333564466904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,512,0.005978666659858491
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,1024,0.0052684446175893145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,128,0.0030337776988744736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,256,0.004596444467703502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,64,0.0028533333291610083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,128,0.0057813334796163775
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,768,0.006292444550328785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,32,0.002838222309947014
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,768,0.005311111195219888
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,65536,0.0503751105732388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,64,0.00462666650613149
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,3072,0.008607111043400234
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,65536,0.055303110016716846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,65536,0.05544177691141764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,16384,0.01642311116059621
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,51200,0.044848889112472534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,256,0.005967999911970562
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,12288,0.013174222575293647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,16384,0.018230222993426852
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,16384,0.018425777554512024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,10240,0.012169777519173093
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,12288,0.014469333820872836
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,128,0.004589333302444882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,12288,0.015156444576051501
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,10240,0.012468444804350534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,10240,0.013408888545301227
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,51200,0.04016444418165419
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,7168,0.010627555350462595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,8192,0.011857777833938599
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,6144,0.009729777773221334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,7168,0.01017955525053872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,7168,0.01071733319097095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,5120,0.008944888909657797
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,6144,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,6144,0.008722667064931657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,4096,0.01057866629627016
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,5120,0.00904177791542477
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,5120,0.008341333104504479
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,4096,0.008332444561852349
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,4096,0.007351111206743453
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,32,0.004596444467703502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,8192,0.011064888702498542
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,3072,0.006969778074158563
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,3584,0.008032889001899296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,51200,0.044713778628243335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,3072,0.006627555522653792
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,2560,0.00738933351304796
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,2560,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,8192,0.01107377807299296
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,3584,0.009722666607962714
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,1536,0.006615110983451207
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,2048,0.007323555648326874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,2048,0.006045333213276333
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,1024,0.0052720002002186244
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,1536,0.006868444383144379
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,1536,0.005633777628342311
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,768,0.004449777719047335
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,1024,0.0063573333124319715
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,3072,0.007653333246707916
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,512,0.0038986665507157645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,768,0.00627377794848548
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,768,0.004968889057636261
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,2048,0.007687110867765214
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,512,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,512,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,1024,0.004915555732117759
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,256,0.004634666774008009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,128,0.005771555420425203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,128,0.004854222138722737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,2560,0.006001777946949005
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,32,0.00291377781993813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,64,0.004581333448489507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,65536,0.03748355640305413
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,128,0.0032133333798911837
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,65536,0.05537511242760552
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,65536,0.054758220911026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,64,0.0028560000161329904
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,3584,0.007024889190991719
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,1024,0.00527022240890397
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,16384,0.013798221945762634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,51200,0.04464711083306206
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,51200,0.04054844379425049
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,256,0.0032426667296224167
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,16384,0.017584888471497428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,256,0.0057751110030545135
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,10240,0.01036088913679123
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,12288,0.014479999740918478
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,10240,0.012091555529170565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,10240,0.011980444192886353
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,8192,0.01015822258260515
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,8192,0.010729778144094678
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,7168,0.009493333597977957
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,32,0.004621333132187526
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,7168,0.010431110858917236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,12288,0.013462222284740873
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,51200,0.030569778548346624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,6144,0.00867911097076204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,5120,0.007643555601437886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,6144,0.009551111194822524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,5120,0.008902221918106079
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,12288,0.011327111058764987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,5120,0.007992000215583378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,3584,0.007106666763623555
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,4096,0.008344888687133789
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,6144,0.009054222040706212
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,4096,0.007294221884674496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,3584,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,3584,0.0069848886794514125
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,3072,0.007648888561460707
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,3072,0.006640000061856375
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,2560,0.007679111427730984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,2560,0.006331555545330048
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,16384,0.016912889149453905
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,4096,0.00742222203148736
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,1536,0.0052533331844541765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,2048,0.0070293330483966405
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,7168,0.009029332962301042
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,3072,0.006646222124497096
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,1536,0.007144889069928064
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,1536,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,768,0.0039022221333450745
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,1024,0.006311111152172089
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,2560,0.005967111223273807
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,1024,0.0052986666560173035
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,768,0.006285333385070165
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,768,0.0052666668262746595
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,256,0.0032071111102898917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,512,0.004971555537647671
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,256,0.005718222094906702
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,2048,0.005948444621430502
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,1024,0.004562666846646203
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,128,0.005607999861240387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,128,0.004612444589535396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,512,0.0035751110149754416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,32,0.0028568889117903183
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,64,0.004879999905824661
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,65536,0.026455110973782007
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,32,0.004611555486917496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,128,0.0028968888024489083
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,51200,0.022336888644430373
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,64,0.002867555452717675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,51200,0.0373946660094791
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,16384,0.014831110835075378
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,512,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,16384,0.01590755581855774
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,12288,0.012772444221708508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,256,0.004931555440028508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,65536,0.04644088943799337
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,10240,0.011559999651379056
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,12288,0.012085333466529846
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,10240,0.012082666986518435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,10240,0.010910222099887
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,8192,0.010371555884679159
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,2048,0.006616888774765863
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,7168,0.009392889009581672
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,8192,0.010716444088353051
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,16384,0.015549333559142219
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,6144,0.00908622228437
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,7168,0.009014222357008193
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,12288,0.013105777402718862
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,6144,0.009672000176376766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,6144,0.008377778033415476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,5120,0.00904977818330129
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,4096,0.007339555356237624
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,8192,0.00962933318482505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,4096,0.008338666624493068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,3584,0.006684444430801604
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,7168,0.010276444256305695
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,3584,0.007995555798212687
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,3072,0.006265777680608962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,5120,0.007997333175606197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,3072,0.007698666718271043
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,51200,0.044072889619403414
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,3072,0.006779555645253923
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,2560,0.007311999797821045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,8192,0.009710222482681274
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,4096,0.007278222176763747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,65536,0.05437511205673218
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,2048,0.007095110913117726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,1536,0.006660444455014334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,2560,0.0057502223385704895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,1536,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,1024,0.0052693333062860705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,1536,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,2560,0.006313777632183499
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,1024,0.006308444258239534
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,1024,0.005221333354711533
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,768,0.005979555762476391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,3584,0.006986666884687211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,768,0.005277333160241445
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,512,0.005945777727497949
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,256,0.003206222214632564
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,2048,0.006960888703664144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,256,0.0046364445653226645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,128,0.003152888889114062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,256,0.005610666755172942
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,2048,0.005961777849329843
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,128,0.005640000104904175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,128,0.004577777865860197
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,32,0.002534222271707323
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,768,65536,0.021439111895031396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,64,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,32,0.0047644442982143825
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,768,51200,0.01909244391653273
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,768,0.004565333326657613
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,51200,0.04313866628540886
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,51200,0.03698666559325324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,64,0.0025537777692079544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,16384,0.01590044465329912
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,512,0.004928888960017098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,768,12288,0.011503111157152386
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,12288,0.01348533315791024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,768,10240,0.010424888796276517
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,768,16384,0.010018666585286459
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,10240,0.010759111079904767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,768,8192,0.009073778159088558
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,10240,0.012076444096035428
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,16384,0.014515555567211576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,8192,0.010666666759385003
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,768,7168,0.008366222182909647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,65536,0.054089777999454074
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,7168,0.010088000032636855
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,768,6144,0.007984889050324758
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,6144,0.009704888694816166
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,768,5120,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,5120,0.007975999679830339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,5120,0.007981333467695449
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,768,4096,0.007289778027269576
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,5120,0.009011555876996782
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,12288,0.011744889120260874
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,7168,0.009019555317031013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,768,3584,0.0069777775141927935
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,3584,0.008009778128729926
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,6144,0.00869511150651508
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,768,3072,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,3584,0.0069724445541699724
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,768,2560,0.006274666637182236
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,3072,0.007661333514584436
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,3072,0.0068693334857622785
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,768,2048,0.004967111266321606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,2560,0.006304889089531369
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,65536,0.0458222230275472
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,2048,0.006105777704053455
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,512,0.004051555362012652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,1536,0.005664888769388199
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,768,1024,0.004286222159862518
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,2560,0.007307555940416124
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,1024,0.0053288886944452924
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,768,768,0.0038977778620190094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,1024,0.006281777802440856
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,2048,0.006633777585294511
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,4096,0.007290666302045186
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,768,512,0.0035422220826148987
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,768,0.005251555393139522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,512,0.004907555464241239
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,1536,0.006692444284756978
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,256,0.00556355549229516
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,768,128,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,4096,0.008349333372380998
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,768,1536,0.0052586665583981415
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,768,64,0.002540444541308615
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,512,0.005599110904667113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,768,256,0.0029208889851967492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,64,0.004755555755562252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,512,65536,0.017103999853134155
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,8192,0.00958577791849772
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,65536,0.04506133331192864
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,512,51200,0.01497511068979899
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,768,0.006115555349323485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,51200,0.038712001509136625
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,512,16384,0.011396444506115384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,256,0.004634666774008009
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,16384,0.014184888866212634
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,512,12288,0.008671110702885522
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,16384,0.015832000308566622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,128,0.004576000074545543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,128,0.005642666584915585
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,12288,0.011753777662913004
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,768,32,0.002846222163902389
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,10240,0.01072533345884747
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,512,8192,0.007988444632954067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,32,0.0046364445653226645
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,65536,0.047919111119376294
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,8192,0.009473777479595607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,512,7168,0.007695111135641734
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,12288,0.01310400002532535
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,512,6144,0.007311111523045435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,51200,0.036637332704332136
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,10240,0.011911999848153857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,6144,0.008004444340864817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,8192,0.0107351111041175
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,5120,0.009034666750166152
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,7168,0.010058666268984476
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,7168,0.008996444443861643
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,4096,0.008339555727110969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,4096,0.0070426662762959795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,512,5120,0.0070222218831380205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,512,3072,0.006600889066855113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,3584,0.007711999946170383
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,3584,0.006638222270541721
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,5120,0.007320000065697565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,512,2560,0.0063164445261160536
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,3072,0.006656888872385025
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,512,2048,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,512,4096,0.00737955586777793
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,2560,0.0059573331640826324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,512,1536,0.004602666530344221
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,2048,0.007279111279381647
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,2048,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,512,3584,0.0069573331210348345
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,512,1024,0.003921777837806278
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,1536,0.0053004444473319584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,512,768,0.003585777762863371
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,1024,0.006280000011126201
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,1024,0.004943999979231092
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,512,512,0.0033173333439562055
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,768,0.004927111168702443
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,2560,0.007359999749395583
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,512,256,0.0032239999208185407
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,512,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,512,0.004589333302444882
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,512,128,0.002858666703104973
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,256,0.005605333381228977
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,6144,0.009713778065310584
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,512,64,0.002875555513633622
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,128,0.005625777774386936
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,512,32,0.00286311118139161
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,64,0.004595555365085602
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,3072,0.007343110938866933
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,256,65536,0.015881778465376962
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,32,0.00423911131090588
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,512,10240,0.007975111405054728
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,65536,0.04636177751753065
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,65536,0.04206933246718513
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,256,16384,0.008314666648705801
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,51200,0.034295999341540866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,256,0.004629333400064045
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,256,12288,0.008027555214034187
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,128,0.004556444370084339
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,16384,0.014134221606784396
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,256,10240,0.007320889168315464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,12288,0.013352889153692456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,768,0.005991111199061076
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,256,8192,0.006973333656787872
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,10240,0.012091555529170565
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,256,51200,0.01310488912794325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,10240,0.010660444696744284
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,8192,0.01072266697883606
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,8192,0.009531555904282464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,51200,0.037363555696275495
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,7168,0.010032888915803697
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,7168,0.008669333325492011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,256,5120,0.005612444546487596
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,6144,0.009531555904282464
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,6144,0.008681777450773451
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,256,4096,0.005372444374693765
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,5120,0.008978666530715095
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,12288,0.011904000408119626
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,256,3584,0.005919111271699269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,4096,0.008350222474998897
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,256,7168,0.006631999793979857
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,256,3072,0.005952000204059813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,3584,0.006656000183688269
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,256,6144,0.006612444503439798
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,256,2560,0.006278222219811545
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,1536,0.006730666591061487
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,3072,0.006303999986913469
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,16384,0.015895111693276297
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,2560,0.007376888559924231
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,2560,0.00629688882165485
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,256,1536,0.00425511101881663
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,2048,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,4096,0.007325333025720384
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,1536,0.006620444357395172
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,1536,0.005600888695981767
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,3072,0.007631999750932057
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,256,2048,0.005616000129116907
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,256,768,0.003556444413132138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,2048,0.007240888973077138
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,768,0.004912888838185204
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,256,512,0.0032071111102898917
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,3584,0.007968000239796108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,512,0.005952888892756567
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,512,0.004592888885074192
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,256,256,0.002867555452717675
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,256,0.005602666487296422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,1024,0.006317333214812809
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,256,128,0.002558222247494592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,5120,0.007359111474619971
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,128,0.005608888963858287
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,768,0.005987555616431766
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,256,32,0.0028764444092909494
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,128,65536,0.014212444424629211
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,64,0.004584889031118817
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,32,0.004567999806669024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,128,51200,0.01275644451379776
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,65536,0.04154933492342631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,256,0.004552888787455029
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,51200,0.037414222955703735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,51200,0.03357866737577651
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,256,64,0.002510222295920054
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,128,12288,0.007087111473083496
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,128,0.004622222234805425
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,16384,0.014159111513031853
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,256,1024,0.003721777763631609
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,128,10240,0.006999111009968652
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,12288,0.013183111117945777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,12288,0.011718221836619906
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,10240,0.012080888781282636
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,1024,0.005256888767083486
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,10240,0.01074577785200543
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,8192,0.01071200023094813
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,8192,0.009374221993817223
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,128,6144,0.005946666830115848
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,128,16384,0.007671111159854465
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,7168,0.00903911143541336
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,128,5120,0.005964444329341252
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,6144,0.009367110828558603
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,16384,0.015855110353893705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,6144,0.007998222278224098
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,5120,0.008674666285514832
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,5120,0.007994666695594788
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,128,8192,0.006660444455014334
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,4096,0.007985778152942657
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,4096,0.00675999994079272
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,128,3072,0.005603555589914322
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,3584,0.007649777664078607
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,3584,0.00665244460105896
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,128,2560,0.005963555640644497
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,7168,0.010096000300513374
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,3072,0.00600622221827507
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,128,2048,0.0053093334039052325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,2560,0.007340444458855524
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,2560,0.005984000033802456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,128,1536,0.004593777573770947
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,2048,0.0070222218831380205
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,2048,0.005613333235184352
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,128,1024,0.0038906666967603895
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,1536,0.006865777903132969
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,65536,0.04625066783693102
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,1536,0.005347555710209741
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,1024,0.0063279999627007385
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,1024,0.004919111314747068
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,3072,0.007300444775157505
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,768,0.004953777624501122
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,512,0.0059164443777667145
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,128,4096,0.005939555664857228
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,128,256,0.0028497777465316984
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,256,0.00564444437623024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,128,128,0.002830222249031067
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,128,768,0.003546666767862108
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,128,0.005617777920431561
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,128,512,0.0031848889258172777
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,512,0.004578666554556953
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,128,32,0.0026319999661710528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,128,3584,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,256,0.004916444420814514
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,64,65536,0.010383110907342698
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,64,51200,0.009368889033794403
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,768,0.006269333263238271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,65536,0.041215999258889094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,64,16384,0.007342221836249034
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,64,12288,0.006495999793211619
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,64,0.004566222015354368
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,16384,0.013920888304710388
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,12288,0.011794666449228922
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,64,8192,0.006680888848172293
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,32,0.00424888895617591
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,64,7168,0.005969777703285217
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,8192,0.009406222237481011
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,128,64,0.002529777793420686
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,7168,0.008676444490750631
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,51200,0.03331466515858968
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,64,5120,0.005621333503060871
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,64,4096,0.005592888842026393
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,6144,0.008042666647169325
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,64,10240,0.005600000007285013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,64,3584,0.005619555711746216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,5120,0.007704888780911763
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,64,3072,0.005977777971161737
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,4096,0.006680000159475539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,64,2560,0.006288888967699475
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,3072,0.006488889041874144
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,64,2048,0.005950222412745158
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,2560,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,64,1536,0.00423200014564726
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,2048,0.005943111247486538
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,64,1024,0.003906666818592283
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,1536,0.0052871112194326185
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,64,768,0.0035751110149754416
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,1024,0.0052488889131281115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,64,512,0.0032151111712058387
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,3584,0.006384000182151794
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,64,6144,0.006272000157170826
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,64,256,0.0028684443483750024
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,768,0.004935111022657818
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,512,0.004911111046870549
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,64,64,0.002530666689078013
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,256,0.00498933345079422
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,64,32,0.002558222247494592
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,64,0.004625777817434735
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,128,7168,0.006600889066855113
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,32,65536,0.009272889130645329
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,32,51200,0.007990222010347579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,32,16384,0.006269333263238271
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,65536,0.041152000427246094
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,64,128,0.002566222308410539
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,32,12288,0.006629333313968446
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,32,10240,0.0063182223174307085
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,16384,0.013986666997273764
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,32,8192,0.005623111294375525
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,128,0.00452622233165635
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,32,7168,0.006630222416586346
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,10240,0.010658666491508484
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,8192,0.009331555830107795
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,32,6144,0.006270222365856171
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,32,5120,0.00629333323902554
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,7168,0.008661333057615492
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,128,0.004646222210592694
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,6144,0.008053333395057255
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,32,3584,0.005619555711746216
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,5120,0.007692444655630324
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,32,3072,0.005922666854328579
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,4096,0.0069617778062820435
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,32,2560,0.005979555762476391
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,12288,0.011423110961914062
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,3584,0.00665155549844106
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,32,2048,0.005249777601824866
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,3072,0.006294222341643439
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,2560,0.005984000033802456
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,32,1024,0.003879111260175705
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,1536,0.005584888988071018
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,32,768,0.003536888708670934
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,51200,0.03326755430963304
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,32,512,0.0032017777363459268
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,768,0.004948444250557157
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,32,256,0.002903999967707528
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,10240,0.010381333529949188
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,512,0.004951111144489712
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,32,128,0.002850666642189026
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,32,64,0.0028524444335036804
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,32,0.0046880000995265115
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,256,0.004572444491916233
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,128,0.004583111239804162
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,32,4096,0.005606222069925732
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,64,0.004582222137186262
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,32,0.004597333156400257
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,32,1536,0.004219555606444676
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,float16,1,32,32,0.0025537777692079544
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,1024,0.0052435555391841466
VLLM,0.16.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,2048,0.005611555443869696
